PMP module book final.pdf - Blackboard - University of Leicester

PROFESSIONAL DIPLOMA IN MANAGEMENT 

School of Management 

Professional Management Project 

MN4507/D 

www.le.ac.uk/ulsm/

school of management 

MODULE MN4507/D 

Professional Management 

Project

school of management

Professional DiPloma in management 


contents 

introduction to the module ....................................................................................... i 

structure of the module ............................................................................................ i 

learning outcomes ................................................................................................. iv 

how to approach this module ................................................................................. v 

Section 1: Research and the Professional Management Project .....................1 


introduction 

What is research? 

Why Do We ask You to Do a Project as Part of Your Diploma Programme? 

What should a Project include? 

What is a good Professional management Project? 

summary 

references 

Section 2: The Role of Existing Literature in Social Science Research ...........11 

introduction 

Why Do You need to Do a literature review? 

additional guidance on finding subject-specific literature 

Using the literature 

compiling a literature review 

summary 

references 

Section 3: Formulating Research Questions ..................................................25 

introduction 

research topic versus research Questions 

What to think about in topic and Question selection 

summary 

references 

Section 4: Quantitative and Qualitative Methods – A Selective Review ........33 

introduction 

Distinguishing Between Quantitative and Qualitative methods 

Quantitative methods 

Quantitative/Qualitative methods 

Qualitative methods

choosing a method 

a note about Using secondary Data 

and finally … 

summary 

references 

Section 5: Sampling, Design and Administration ..........................................51 

introduction 

sampling 

Designing research schedules 

Designing schedules for methods Using Direct Questions 

Designing self-administered Questionnaires and structured interview schedules 

Designing semi-structured and Unstructured interview schedules 

Designing observation schedules 

administering research schedules 

summary 

references 

Section 6: Planning, Access and Ethics ..........................................................77 

introduction 

research Planning 

research access 

research ethics: the Basics 

maintaining confidentiality and anonymity 

tensions and ambiguities in research ethics 

summary 

references 

Section 7: Analysing Quantitative Data .........................................................97 

introduction 

the terminology of statistics 

Descriptive Data 

representing Descriptive statistics 

statistical inference 

summary 

references 

Section 8: Analysing Qualitative Data ..........................................................115 

introduction 

Data analysis in general: a Brief overview 

Deductive analysis of Qualitative Data 

inductive analysis of Qualitative Data 

content analysis 


school t of management 

software Packages and analysing Qualitative Data 

Presenting Qualitative Data 

some final Pointers on the analysis of Qualitative Data 

summary 

references 

Section 9: Writing Up Your Professional Management Project ...................133 

introduction 

the importance of good Writing in Your Project, and how to achieve it 

some technical Pointers, or the ‘Dos’ and ‘Don’t’s’ of Writing a Project 

other sources of advice and information on Writing Your Project 

summary 

references 

Appendix 1: Doing a Theoretical Project .........................................................147 

Symbols Used in this Module 

� Key Reading 

� Tasks 

this module is supported by a companion textbook. Details of 

which chapters or parts of the textbook you need to read in 

conjunction with each section of the module book are provided on 

Blackboard. these are your key readings. You should complete the 

appropriate key reading where this symbol appears and directs you 

to Blackboard. 

the module also contains a series of tasks. these indicate 

opportunities for connecting module concepts with your own 

studies or practical organisational situations. again these will be 

provided on Blackboard. You should complete the appropriate task/s 

where this symbol appears and directs you to Blackboard.


Professional management Project 


Introduction to the Module 

Welcome to the module book for the Professional Management Project, the last 

module in your Professional Diploma in management programme. congratulations 

on your progress so far! 

this module book 1 deals with what we call research methodology, because it is 

designed to assist you in undertaking independent research for the Professional 

management Project (hereafter just ‘project’, or ‘PMP’), which is also the module 

which carries the most credits in the Diploma programme – 30, to be precise! so 

what is research methodology? the straightforward answer is that it is an overall 

action plan for research. in the social sciences 2 , methodology is the logic or series 

of steps that connects a given set of research questions (uncertainties or gaps in 

our knowledge about the social world, about human behaviour) to the conclusions 

arrived at. it encompasses the selection of research methods, the design of data 

gathering instruments like interview schedules or self-administered questionnaire 

schedules, gaining access to the research site, sampling, research ethics and data 

analysis. academic standards exist on what constitutes good research. this module 

is intended to assist you in familiarising yourself with these standards so you are able 

to justify your methodological choices to your assessors, gather meaningful, relevant, 

credible, trustworthy and plausible data which will answer your research questions 

and undertake your project research in an ethical way. 

not every aspect of this module will be relevant to every Diploma student when 

preparing their projects, as all projects are different. so you will have to decide for 

yourselves which bits you need for your own research. nonetheless, you should read 

the module book in its entirety, as well as completing the tasks where instructed to, 

in order to familiarise yourself with methodology as a subject area. this should help 

you to make the appropriate choices for your own PmP. in addition, please be aware 

that cross-references within and across the sections will be made as appropriate, and 

that material may be initially introduced and then developed in more detail at a later 

stage according to the structure of the module itself. 

1 as with the other modules, key terms appear in bold in this book when they 

are first used. 

2 Natural sciences study the behaviour of natural organisms like plants, animals, 

natural forces like gravity and natural elements like iron or silver – they include biology, 

physics and chemistry. the social sciences study human behaviour, and therefore 

management is a social science. Please also note that the term ‘management’ is used 

in this module book as a generic term to refer to the collection of subjects, theories, 

techniques and concepts which you have studied during the previous six modules on 

the Diploma programme (whether these cover people’s behaviour at work, strategy 

formation and implementation, accounting, finance, marketing, statistics etc.). 

school of management i


there are many and varied methods, ways of gaining research access, sampling 

techniques, data analysis tools, ethical issues and debates and so on in the social 

sciences and it would be impossible to cover all of these here. instead we will be 

exploring issues related to the ones which, based on Ulsm experience, most students 

of management use or need to be aware of. for those of you who do not find the 

specific aspect of methodology which you are interested in covered here, this will be 

discussed in detail in a range of the available texts. Key reading is as usual stipulated 

for this module (more details below). You can also find references at the end of each 

section in this book and a suggested additional reading list on Blackboard. 

Structure of the Module 

the module consists of nine interrelated sections. 

section 1 offers an introduction to research and the PmP. this section provides 

guidance on the nature of academic research and what Ulsm expects from Diploma 

students in their project submissions. 

section 2 discusses the role of a literature review in research. it also offers advice on 

how best to develop your literature review. 

section 3 covers the formulation of research questions. it outlines the various things 

you need to bear in mind when choosing and refining your area of academic interest 

and your research questions, so that you can work towards identifying clear questions 

as well as identifying plausible reasons as to why these questions are important. 

section 4 discusses research methods. this section provides an overview of the ones 

most students of management use, as well as offering advice on how to choose a 

method(/s). 

section 5 covers sampling for, designing and administering research schedules. 

in this section we discuss how to select your respondents (sampling) and how to 

design your interview or questionnaire or observation schedule. We also offer some 

tips on how to conduct the data gathering itself, depending on which method you 

choose (research administration). 

section 6 discusses how best to manage your time when doing research (planning), 

and also covers the sometimes challenging issue of obtaining research access (actually 

getting into an organisation, for example, to collect data). the last issue in this section 

is research ethics, which we take very seriously indeed. You need to do the same! 

section 7 is an overview of the analysis of quantitative data. here we identify the 

different types of statistics and different types of quantitative data. We also discuss 

various ways in which quantitative data can be measured, and cover some key 

techniques for representing descriptive statistics in graphical form and in numerical 

form. finally this section discusses the role of inference in statistical analysis 

ii school of management


section 8 is an overview of the analysis of qualitative data. Qualitative data are 

expressed in words, and usually collected in response to open questions or a nonstructured 

observation schedule. they are often considered to provide a more indepth 

and richer account of a research site. this section discusses some of the choices 

available when analysing qualitative data, and some of the key issues to bear in mind. 

section 9 discusses the writing up of your project. since the mark you receive for the 

project depends entirely on your written submission, this section offers help on how 

to produce the best piece of writing you can in this regard. 

as many examples as possible will be given to illustrate how the issues raised are 

visible in real-life social science/management research. the module is arranged in a 

logical order according to how research is usually carried out in practice. it is therefore 

sensible to begin at the beginning and read all the way through to develop a gradual 

awareness of the process of academic research and the tasks involved. 

methodology as understood here, then, has to do with the practicalities of putting 

together an empirical piece of research. ‘empirical’ comes from the greek empeiria, 

meaning “knowledge based on experience and observation” (gummesson 2000:64). 

in the social sciences, it refers to research which uses real-world data, which bases its 

conclusions on findings about people’s actual behaviour and activities, as opposed to 

engaging in abstract theorising, reflection and conceptualisation. in practice though 

it is not always easy to draw the line between empirical projects (which can involve the 

analysis of secondary data – which someone else has collected – as well as primary 

data – which the researcher collects themselves) and what are sometimes referred to 

as theoretical, library or desk-based projects, where the researcher focuses on an 

evaluation or synthesis (bringing together) of the existing literature in a particular 

subject area/s. this is because theoretical projects often involve the analysis of existing 

empirical studies – and thus the use of secondary data. and we should not forget 

that the analysis of any form of data, whether primary or secondary, should always 

involve reference to existing concepts and theories – so empirical projects also need 

to make considerable use of theory! 

in this book, however, we will focus on primary empirical research for the following 

reasons: 

a) again, Ulsm experience shows that the majority of students of management 

undertake empirical studies for their projects, dissertations or theses, usually 

involving the collection of primary data of some sort; 

b) even if you choose to engage in a ‘pure’ library project or to analyse secondary 

data only for your project, you may well be asked to conduct primary empirical 

research in the future, either as part of future academic study or in your 

working life; 

c) a grounding in primary empirical research methodology is always helpful in 

terms of evaluating claims made by others on the basis of their data-gathering 

activities (i.e., in terms of evaluating published management research); and 

school of management iii


d) it is impossible to cover the full range of available methodologies, as already 

suggested, in a module like this. 

moreover, many of the issues (e.g., developing a literature review, formulating 

research questions, data analysis and writing up your project) that we discuss during 

the module will in any case be directly relevant to those doing more theoretical 

projects and/or projects involving the use of secondary data. however, Appendix 1 

offers some guidance to those of you who choose to do a theoretical project. 

Learning Outcomes 

this module, as already suggested, will not tell you all you need to know about your 

own project research – so it is essential that you do as much independent reading as 

possible. You will also need to read a wide range of the subject-specific literature of 

course – i.e., published research on your area of interest, whether it is organisational 

structure, organisational culture, relationship marketing, charismatic leadership, 

stress, productivity, employee motivation, the supply chain, auditing or whatever else. 

moreover, a familiarity with the methodological literature beyond the key readings 

prescribed here will only enhance the decisions that you make in this regard for your 

project, and the finished product itself. 

Please also be aware that terminology in the area of research methodology, like many 

other areas of the study of management, is not used consistently. so authors might 

use different terms to describe the same thing or (even more confusingly) the same 

term to describe different things. Be aware of this when you are reading this book 

and the relevant texts. 

more formally, at the end of this module, typical students should be able to: 

1. formulate and undertake a piece of original research with relevance to 

contemporary management issues and problems 

2. integrate and interrelate concepts, techniques and skills acquired throughout 

the course of the Professional Diploma programme 

3. demonstrate an appreciation of relevant existing research and theoretical 

perspectives relating to the project topic 

4. develop and apply analytical and communication skills needed to accomplish 

the project process. 

iV school of management


How to Approach This Module 

in designing and writing this module book, we are aware that students will approach 

their PmP from a range of different backgrounds and will bring to this research a 

varied assortment of prior knowledge and expectations. 

to help you to realise the learning outcomes, and as with previous modules on the 

Diploma programme, this module comprises the following components: 

1. this module book 

2. the companion textbook 

3. the virtual learning environment (Blackboard). 

to successfully navigate the module and to get a thorough grasp on the subject 

matter you will need to use the module book, companion textbook and Blackboard. 

as suggested above, this module book provides guidance on: what research is and 

what we expect of your PmP; the functions of a literature review in research and 

how to compile one; the development of appropriate research questions; the choice 

of a research method; sampling, schedule design and administration; the analysis 

of quantitative and qualitative data; and writing up the PmP. the authors are Ulsm 

academics with considerable experience of supervising management students and 

conducting their own research, 

to reiterate, you should work through the book in order. as you do so, you should try 

to think about the information and advice it offers in relation to the kind of project 

you would like to do. as well as all sorts of examples from real research projects, the 

book also includes some thinking points which ask you to reflect on a particular 

section of content using some stimulus material. 

further, the module is supported by a companion textbook. all key readings from this 

text, which is provided, are indicated on the Blackboard site for this module. Blackboard 

also contains some suggested additional readings. again as already stated, a list of 

references for the material that has been used to compile the module book will be 

provided at the end of each section here. You might also want to consult these for 

further reading. as ever, you will find that there is more additional reading available 

with this module than you can feasibly read, so you will need to be selective in your 

choice of extra reading. Please also do not feel restricted to the readings suggested 

– reading around the subject area is encouraged. if you do find readings that you 

feel others would benefit from, please share them on Blackboard by providing the 

hyperlink and/or citation. 

and so to the module Blackboard site. in addition to additional readings, the site 

features tasks for each section of the module in order to develop your understanding 

of how to do academic research. You will also find streamed lectures, links to relevant 

school of management v


websites and several support forums there. the support forums – which are divided 

by subject area – allow you to discuss the tasks in this module with others who are 

progressing through the module at the same rate as you are. also, and importantly, 

the tutor on the support forum for the subject area you choose to research also acts 

as your project supervisor. You should therefore make use of this forum as much as 

you can to make sure you are going in the right direction. the Blackboard site also 

contains the Professional Management Project Guidelines, which you should read 

in full. these guidelines cover issues such as what to expect from the supervision via 

the support forum, how to format your project and so on. 

finally, all the references and Urls in this module book were correct at the time 

of going to press. We will provide any updates of which we become aware on the 

Blackboard module site. 

References 

the following are the sources which were used to compile this section. chapters 

or excerpts are specified to suggest material which should be especially relevant to 

issues covered in the section. 

gummesson, e. (2000) Qualitative Methods in Management Research 2 nd edition 

thousand oaks, california: sage chapter 3. 

vi school of management



section 1 

Research and the Professional 

Management Project



PROFESSIONAL MANAGEMENT PROJECT 

SECTION 1 

Research and the Professional 

Management Project 


having studied this section of the module, you will be able to: 

• define what research means in an academic context 

• understand why we ask you to conduct your own research project 

as a mandatory part of the Professional Diploma in Management 

• identify the key components of a Professional Management 

Project 

• outline the features of what ULSM considers to be a good project. 

Introduction 

As established, this module is intended to support you in the process of putting 

together your Professional Management Project. This independent research project is 

a mandatory part of your Diploma programme. As we stated in the introduction, it is 

also a particularly crucial part of this programme, as it gains you 30 credits out of a 

total of 120! So let’s begin with a consideration of what the term ‘research’ means. 

What is Research? 

Williams and May (1996:7) suggest that “Research may be characterized as methodical 

investigations into a subject or problem. To ‘research’ is to seek answers that involve 

understanding and explanation.” In other words, research is about answering 

questions, seeking answers to things we are not sure about, clearing up gaps in the 

breadth or certainty of our knowledge about the social world. The term ‘research’ 

actually implies uncertainty – re-search – look again. Research is usually not just 

descriptive – i.e., it doesn’t just seek to provide an accurate portrayal of what is 

going on in a specific situation. Instead it tends to be explanatory – that is, looking 

for reasons as to why particular things are happening, aiming to establish what the 

cause and effect relationships are in this specific situation. Or it can be exploratory 

– seeking insights into or asking questions about unfamiliar or complex situations, or 

SChOOL OF MANAGEMENT 1


trying to see a situation from a different, unconventional point of view. So analysis 

of research data usually produces insights, explanations, relationships, comparisons, 

predictions or theories. 

Equally, “Much enquiry in the real world is essentially some form of evaluation” 

(Robson 2002:6). So empirical research tends to try to understand what is going on 

in the ‘real world’ and why, but also to make a judgement about it. Management 

research especially is often directed at solving a particular organisational problem, 

or improving organisational processes in specific ways. As such the knowledge 

that derives from this research is applied to the ‘real world’. For you as students 

of management, then, the emphasis of your project may well be on identifying a 

management or business problem, formulating solutions and generating appropriate 

and acceptable recommendations for action. however, as we have seen, it is also 

possible to undertake a more theoretical piece of work that focuses in detail on an 

issue in management thought. 

But why do you have to do your own independent research project in order to gain 

a Professional Diploma in Management qualification from the University of Leicester 

School of Management? 

Why Do We Ask You To Do a Project as Part of 

Your Diploma Programme? 

There are many reasons for asking you to do project research. First, we believe that 

undertaking this kind of project sharpens your information gathering, critical and 

analytical skills. Second, it enhances your subject-specific knowledge. Note that this 

is not just for academic reasons – in other words, we don’t believe that you should 

know more about specific management subjects just for knowledge’s sake. Instead we 

believe – as we hope has become obvious by now! – that better awareness of debates 

and issues in academic literature about management will assist you to underpin your 

work practice with solid reasoning and informed reflection, and therefore improve 

this practice. Project research also often requires you to relate academic theories and 

concepts to real-world data and real-world problems. In other words, a project may 

well allow you to identify the gaps between management theory and management 

practice and suggest why these gaps exist – is it the practice which is problematic or 

the theory, or a mixture of both? This may also challenge your preconceptions about 

effective management practice. 

Fourth, undertaking a project means that you have to identify significant issues and 

themes yourself – finding, sorting through and organising a large body of academic 

literature and, in all likelihood, empirical data to pull out what the important issues 

and themes are in terms of your research questions. Project research also develops 

your interpersonal/transferable skills – including oral and written communication, 

negotiation, problem solving, time management, self-motivation and creativity. And, 

finally, you have to take responsibility for the whole process. Doing a project means 

that you make the choices from start to finish – you set the question/s, organise the 

2 SChOOL OF MANAGEMENT


process, decide on the relevant literature, consider and design your methodology, 

gather the data and reach your own conclusions based on the analysis. The project is 

therefore unique as regards your Diploma studies, as ULSM tutors set the questions 

and the structure for all the other pieces of assessed work that you do. 

OK, so we know now what research is, and why we ask you to do independent 

research as a significant aspect of your Diploma programme. But what does a PMP 

look like? 

What Should a Project Include? 

Your project may not necessarily be organised in the following way because every 

project is different but for empirical projects all these components need to be 

present. Please also note the ULSM administrative requirements in your Project 

Guidelines on Blackboard – e.g., regarding the title page, the contents page/s, the 

acknowledgements, the overall format of the project and so on. 

Executive Summary/Abstract 

This should be no more than one side of A4 in length. It is a self-contained summary of 

the whole of the project. It should contain information about your research questions 

and why these are important; the terms of reference of the project (e.g., how you 

are defining certain key concepts or the parameters of the project); information about 

the subject-specific literature you have used; information about how you answered 

your research questions (i.e., your methodology and your data analysis, if relevant); as 

well as a summary of your conclusions, recommendations and action plan (if any). 

See Saunders et al. (2009:532–533) for extra advice. Even though it appears first, 

you should write the executive summary last of all, so that it actually represents the 

contents of the project! 

Introduction 

The introduction is effectively an extended executive summary and should include 

details of your research questions and why they are significant, your terms of reference, 

the sources of information on which the project is based and how this information 

was collected. The introduction sets the scene and puts the whole enquiry into its 

proper context. We suggest that you structure it as follows – begin with a clear 

definition and justification of the research questions you are investigating, including 

a statement of the management issues involved. You may also wish to acknowledge 

the ways in which these questions changed as the project evolved, if relevant. The 

introduction should then provide an overview of the project as a whole, chapter 

by chapter. Again, even though it forms an early section in the finished product, it 

should be written at the end, for the reasons outlined above. 


Literature Review 


This chapter outlines the academic context of your project and establishes how your 

research adds to or extends what is already known about the topic. That is to say, it 

reviews the existing research in the area – on organisational structure, organisational 

culture, relationship marketing, charismatic leadership, stress, productivity, employee 

motivation, the supply chain, auditing or whatever – identifying the key themes 

and debates, but also identifying the gaps and omissions which your research will 

address. More details of the literature review follow in Section 2. 

Methodology 

This chapter of the project gives details of your data gathering. As already suggested, 

empirical data can be primary – you collect it yourself, for example via questionnaire 

or interview. The methodological information we require about any primary data 

gathering is as follows: 

• discussion of the research method you have chosen and why; 

• discussion of your sample (size and type) and why you chose this size and 

type of sample; 

• discussion of how you negotiated access (if relevant) and why you chose this 

approach; 

• discussion of the design of your schedule and why you did it that way; 

• discussion of your pilot test; 

• discussion of your chosen channel and why you selected it; 

• and discussion of what happened when you actually did the data gathering. 

All these issues are covered in more depth in Sections 4–6. 

But empirical data can also be secondary – i.e., already in existence, collected by 

someone else. So secondary data can come from published academic research, 

government surveys, organisations’ personnel records, shareholder reports, 

newspaper articles and so on. Methodological information needed here would 

include discussion of your data source(/s), including the time period it covers, and 

why you chose it; discussion of how you used this source(/s) and why; and discussion 

of any relevant issues that occurred while you were accessing and using these data. 

Any methodology chapter should also end by discussing the problems that you faced 

during this stage of the project and whether you were able to resolve them; i.e., 

the limitations of the methodology and their implications. You will return to these 

issues in your reflections section. It is also important here, as the above discussion 



indicates, to justify the methodological decisions that you have made throughout 

this chapter. 

Data Analysis 

This chapter should tell the examiners how you did your data analysis, what it shows 

us and how this relates to your research questions and the existing academic literature 

on the subject. More details follow in Sections 7 and 8. 

Conclusions/Reflections/Recommendations 

This chapter draws the project together. You should restate your research questions 

here and summarise the answers to them: these are your conclusions. These 

conclusions must be drawn from the body of evidence presented in the main chapters 

of the project. They should all flow clearly from the preceding analysis. This section 

also needs to identify any problems or opportunities which you have discovered as a 

result of your analysis and which will form the basis of your recommendations. 

In your recommendations you should suggest ways of addressing the problems or 

opportunities discussed in the foregoing analysis. The recommendations chapter might 

also include the benefits of implementing these suggestions but also the resources 

entailed, the action plan/programme of work that would be needed and how long it 

would take. Again the recommendations should be derived from your conclusions. 

You need to be realistic about the extent to which the scope and approach of your 

research allow for firm recommendations to be made, as well as acknowledging any 

limitations or need for further research in this regard. Please also note that your 

project is not a management report or a piece of consultancy. In other words, 

despite the emphasis that we have placed above on its being practically relevant, we 

do not want you only to provide managerial advice. The project should also have clear 

hallmarks of academic rigour, offer potentially generalisable conceptual insights 

and add to or extend the published academic research on the topic in question. 

Your reflections section consists of an overall evaluation of the project – its strengths 

and weaknesses, constraints encountered during the process of putting it together 

and how these were dealt with if appropriate. We would also expect you to reflect 

further on how effective your methodological approach was and to comment on the 

managerial and academic competencies you have developed as a result of writing 

the project. The following questions may be useful in structuring this section: 

• Were your project questions well defined? Did you actually answer them? 

• Did the research outcomes match your initial expectations? 

• Did you do a good job of planning and undertaking the research overall? 

what went well? What should you have done differently? 

• What will you take away from the project process into your management 

career and/or any future studies? 


References/Bibliography and Appendices 


References/bibliography is self-explanatory, but we do require you to provide a 

full list of all the sources you drew on for the project – subject-specific literature, 

research methodology literature and secondary data sources. You should include 

Internet sources here as well. See your Programme handbook and the Foundations 

of Management module book to refresh your memory on how to reference properly. 

Appendices include information which is necessary to the project but would spoil 

its flow or structure if included in the main body. Examples might be a clean version 

of your questionnaire or interview schedule, the letter you wrote requesting research 

access, detailed tables of statistics or graphs relating to your data analysis and so on. 

Do not append raw data (e.g., completed questionnaires or transcripts of interviews) 

here. These should simply be retained in a safe place until your PMP mark has been 

confirmed, for reasons discussed in Section 6. At this point the data can be destroyed, 

if you see fit. 

What is a Good Professional Management Project? 

To expand a little on the basic structure of an empirical project as outlined above, we 

are often asked what constitutes a good project. Well, the answer is one that … 

a) focuses on a contemporary topic which is of practical relevance and personal 

interest to the student. Personal interest in the topic is really important 

because it keeps you motivated and also means that your enthusiasm shines 

through in the finished product; 

b) clearly explains and defends research questions – i.e., you tell us exactly 

what it is you want to find out about your research topic, and you also tell 

us why these questions are interesting. Ideally, your defence should be a 

combination of there being gaps in the existing academic literature, of your 

own personal interest in the topic/questions and of current management and 

academic interest in the topic/questions. See Section 3 for more discussion of 

these issues; 

c) critically reviews as wide a range of subject-specific literature as possible 

and makes its academic contribution clear. Contribution is something we 

expand on as this book progresses, but for now it suffices to say that we 

prefer student research projects to add something to the existing literature 

on organisational structure, organisational culture, relationship marketing, 

charismatic leadership, stress, productivity, employee motivation, supply 

chain, auditing … or whatever you are studying. In other words, a good 

project doesn’t just rehearse or repeat things that we already know. More on 

this in Section 2; 



d) contains a well defended and appropriate methodology. This chapter 

therefore needs to demonstrate all the methodological choices you have 

made but also provide a robust justification for these choices; 

e) clearly and coherently presents and analyses data. More to follow on this 

in Sections 7 and 8, but for now it is important to state that data analysis 

isn’t just about describing the data – i.e., identifying their main themes and 

patterns. It also requires that you analyse/interpret your data – i.e., tell 

the assessors what these themes and patterns suggest about your research 

questions and the existing literature in the subject area; 

f) provides conclusions that follow from the data analysis and feasible 

recommendations that follow from the conclusions. In other words, avoid 

the temptation to conclude or make recommendations on the basis of what 

you think should be happening in this organisation/s, or to make wide 

ranging recommendations that your data don’t really permit. Ask yourself 

whether your data and your analysis actually support these conclusions/ 

recommendations; 

g) offers self-aware reflections. This is one of the sections students often 

neglect or, even worse, use to tell us how utterly perfect and marvellous their 

research is! Make sure that you have spent some time considering what you 

have learnt from the project process and in particular what you would do 

differently if you were to do it again; 

h) includes appendices which support the main text. Remember that really 

important items should be included in the main text rather than requiring 

frequent reference to the appendices, which can irritate readers. Appendices 

should also be mentioned at appropriate points in the text – and do not 

overdo them. The rule of thumb is, if the main text can survive without it, 

leave it out; 

i) accurate referencing throughout and a full list of references at the end. 

This is a matter of academic rigour, avoiding plagiarism and ensuring that we 

can follow up on any sources which interest us at a later date, as you already 

know; 

j) well written, logically structured, carefully presented. Leave enough time 

to do a thorough spell check as well as proofreading for typographical 

mistakes, grammatical errors, logical problems etc; 

k) overall tells a story and is ‘topped and tailed’. Your project should read like 

an unfolding story as follows: 

What are my research questions and why are they important (introduction)? 

What’s the academic background to these questions (literature review)? how 

did I answer them and why did I do it this way (methodology)? What were 

the answers and how did I arrive at them (data analysis)? What does this say 



overall about my research questions? how can this be applied in the real world 

and in general? how good a job did I do (conclusions, recommendations and 

reflections)? 

Topping and tailing refers to presenting each chapter in an introduction/ 

discussion/summary format. So, in each chapter, tell us what you are going 

to do, do it, then remind us what you have done before going on to say what 

the next chapter will do. 

We have now come to the end of our first section, so a short summary is in order 

before moving on to Section 2. 

Summary 

i) Doing research involves systematically asking questions about things we are 

unsure about, which also often implies evaluation of the outcomes. 

ii) Methodology consists of an empirical action plan to enable us to answer 

those questions. 

iii) There are many reasons why we ask you to do an independent research 

project as part of your Diploma studies. 

iv) An empirical project must contain certain components, although not 

necessarily in the order outlined here. 

v) A good project also displays certain key characteristics, perhaps the most 

significant of which is the author’s enthusiasm for the project. 


� Tasks 

You should now visit Blackboard to identify the key reading from 

the companion textbook and to complete the tasks for this section 

of the module. You may also wish to discuss these tasks with the 

tutor and other students on the relevant module support forum, but 

do remember that others may be progressing through the module 

either more slowly or more quickly than you. Additional support 

material for this section can be found in the same section of the 




References 

The following are the sources which were used to compile this section. Chapters 



Blaxter, L., C. hughes and M. Tight (2001) How To Research 2 nd Edition. Buckingham: 

Open University Press chapter 1 

Bryman, A. and E. Bell (2007) Business Research Methods 2 nd Edition. Oxford: Oxford 

University Press pp. 74–77 

Easterby-Smith, M., R. Lowe and A. Thorpe (2008) Management Research 3 rd Edition. 

London: Sage chapters 1 and 2 

Robson, C. (2002) Real World Research: A Resource for Social Scientists and 

Practitioner-Researchers 2 nd Edition. Oxford: Blackwell chapter 1 

Saunders, M., P. Lewis and A. Thornhill (2009) Research Methods for Business Students 

5 th Edition. harlow: Financial Times Prentice hall chapter 1 and pp. 532–533 

Williams, M. and T. May (1996) Introduction to the Philosophy of Social Research 

London: UCL Press chapter 1 






section 2 

The Role of Existing Literature in 

Social Science Research




SECTION 2 

The Role of Existing Literature in Social 

Science Research 


having studied this section of the module, you will be able to: 

• recognise the importance of a literature review 

• understand how to use other people’s research in order to 

determine how and where your work will fit into the current 

subject-specific literature 

• understand how to locate literature relevant to your research 

• outline the key aspects of compiling a literature review. 

Introduction 

Literature reviews are an essential part of any research whatsoever, but this section 

aims first to explain why you need to do one, before moving on to the best way to 

go about accessing relevant sources. We then offer some advice on note-taking and 

subsequently on putting the finished product together. 

Why Do You Need to Do a Literature Review? 

To Refine Your Research Questions 

Obviously by the time you come to read the subject-specific literature you will have 

an idea of your research topic (see Section 3), because without one you wouldn’t 

know which body of literature to access! however, reviewing this literature in 

the first instance also allows you to identify specific concepts, theories, models 

and findings to explore. In other words, what precisely is it about organisational 

structure, organisational culture, relationship marketing, charismatic leadership, 

stress, productivity, employee motivation, the supply chain, auditing or whatever else 

that you want to research? In fact, helpfully, the available literature often explicitly 

suggests directions for future research – especially at the end of journal articles. 


To Contextualise Your Research Topic 


In the finished project, the literature review allows you to inform your assessors about 

how your research fits into what has been previously published on the topic. It should 

outline the existing ideas and themes that you want to focus on in your work, and 

therefore identify relevant academic material which will be used in analysing your 

specific empirical site. So in this regard you are locating your research in the context 

of the existing body of knowledge on the topic, whatever that might be. 

To Enhance/Demonstrate Your Subject-Specific Knowledge 

We have discussed why we see subject-specific knowledge as so important to you 

as students of management in Section 1. however, you also have to demonstrate 

to your assessors that you have a reasonable grasp of the relevant subject-specific 

literature in order to convince them that your research has been carried out with a 

degree of academic rigour, and that it is underpinned with the appropriate level of 

effort and endeavour. 

To Identify Your Contribution 

Related to what we have said above, an effective literature review helps you to avoid 

simply repeating existing studies because it identifies the limitations and gaps in the 

existing literature. Your project should aim to address these limitations and gaps 

either empirically or by evaluating or synthesising literature in new ways. Making a 

contribution means researching something which has not been studied before or has 

not been studied very extensively, adding to or extending what already exists on the 

topic, combining bodies of thought or viewpoints in different ways, researching a 

topic in an unusual way etc. It is a requirement in PhD theses but, to a lesser extent, 

in Master’s dissertations and Diploma projects as well. Original work does not just 

rehearse old arguments to reach already established conclusions. It avoids well worn 

topics and/or uses samples which have not been researched in detail before and/or 

employs unconventional research methods. And it will always be more sympathetically 

received by your assessors! 

To Help You Decide on Your Methodology and Methods 

Reading the available literature in a particular subject area may also help with your 

methodological decisions. This can take two forms – first, there may be a tried and 

tested methodology which is effectively the ‘standard’ in a particular area of research 

and which you can justify using on that basis. Second, and alternatively, it may be 

that most or all researchers in a subject area use one particular methodology but you 

feel you can make a contribution as above, and look at the subject from a new angle 

by choosing a different methodological approach. For example, one of the authors of 

this book decided to use semi-structured interviews in her doctoral research on sexual 

harassment, because once she had read a range of the available literature on this 

topic it became clear that most sexual harassment researchers used self-administered 

questionnaires (SAQs). 



To Ensure Your Research is of Contemporary Interest 

Your literature review also allows you to ensure that your work will be seen as significant 

by today’s management academics and practitioners, if it is done effectively and is 

brought right up to date. You should also make sure that you check on a regular basis 

to see if there have been any significant new publications in your area, especially in 

academic journals, after you have compiled your first good draft of the literature 

review. 

One final point before we move to how you might actually locate relevant literature is 

a clarification of the distinction between a literature review and the use of secondary 

data in empirical studies. As suggested in the introduction and Section 1, some of 

you will be using secondary data to answer your research questions – i.e., data that 

already exist, that someone else has collected. But there is a key difference between 

a literature review and the use of secondary data in projects. A literature review 

summarises the academic material on the subject area in terms of setting the context 

for your project and clarifying your contribution, as discussed above. It therefore 

forms part of the background to your research questions. Which area/s of the 

academic literature are you deriving your inspiration from? Which gaps or omissions 

in this material is your research intended to address? A literature review often makes 

reference to other people’s empirical data and findings in setting this context. It 

should usually precede your methodological discussion and your data analysis, as in 

the conventional project structure laid out in Section 1. 

Data analysis which makes use of secondary data on the other hand deploys such 

data as a direct means to answer your research questions – i.e., as empirical data, 

not as background or context as above. This material should, as also outlined in 

Section 1, follow the methodology and precede the conclusions, recommendations 

and reflections. There are different emphases here in terms of how other people’s 

data might be used in a project. Please do not make the mistake of referring to the 

literature review as a research method. If you are confused about the difference, 

think about how you are using the data in question. Is it simply to set the scene for 

the research you are doing (literature review) or are you actually taking someone 

else’s data and analysing them in a new way (secondary data)? 

To summarise the above points, the important thing to remember is that 

“Knowledge does not exist in a vacuum, and your work only has value in 

relation to other people’s. Your work and your findings will be significant 

only to the extent that they are the same as, or different from, other 

people’s work and findings.” (Jankowicz 2005:162) 

In terms of locating relevant sources for your literature review, we are now going to 

direct you back to the guidance we provided in your Foundations of Management 

module book. Re-read Section 4 on Learning Resources there before you proceed 

with this module book. 



Additional Guidance on Finding Subject-Specific Literature 

Most literature searches will start with a broad and non-selective process which 

Easterby-Smith et al. (2008:34), following Selvin and Stuart, call ‘trawling’: “a wide 

overall review of the literature in a specific field”. This should involve as comprehensive 

an overview as you can manage of the relevant material to ensure breadth of coverage. 

As you proceed through the project and your focus and ideas become clearer, you 

will probably need to ‘fish’ for specific issues, theories etc. about which you want to 

know more. This Easterby-Smith et al. (2008:34, again following Selvin and Stuart) 

explain as follows: “[the researcher] know[s] exactly the articles that they want and 

simply need[s] to collect these”. 

You can start trawling by using ULSM reading lists from the modules you have already 

studied on the Diploma. But you will also need to do your own searches to make 

your review as comprehensive as possible, and therefore to be familiar with relevant 

University of Leicester resources like the online library databases etc. – as of course 

you should be by now. Identification of key authors and seminal contributions is 

especially helpful at the start of the process. For example, a project on organisational 

structure would be expected to make reference to the very important, if controversial, 

work of Max Weber on bureaucracy. Similarly, a project on cross-cultural differences in 

employee behaviour would be expected to note and discuss the (equally controversial) 

work of Geert hofstede in this regard 1 . Review articles also appear from time to time 

in the academic management journals, discussing a particular body of literature and 

the relevant developments in this area – again these are good resources for starting 

your literature review. And of course you can (and should) follow up references from 

texts you have already read by using the bibliographies in those texts. 

If you are having difficulty finding material which is specific to your needs, consider 

broadening your parameters. First, you are free to use literature which has been 

published in languages other than English for your project, as long as it is relevant 

to your research, as well as being a robust source in its own right. Of course you will 

need to translate both the ideas in these texts (including direct quotations) and the 

necessary references into English in the finished product, whilst indicating that these 

are your translations from Greek, Cantonese, German etc. Second, many of you will 

find a lot of useful material in other social science and humanities disciplines – like 

sociology, psychology, media studies, cultural studies, history, philosophy, politics, 

international relations, anthropology etc. In other words, 

“[y]ou should not be limited by the research (and research questions) 

current in the specific field you are researching. Researchers in other 

fields and from other disciplines may well be wrestling with problems 

similar to yours, or from which useful parallels can be drawn.” 

(Robson 2002:57) 

1 Further to this, if an issue or author like this is central to your research, then 

we would expect you to read the relevant original – not just to rely on how other 

authors describe this work. 



Third, as a hypothetical example, say you want to do research on charismatic 

leadership in hotels, but can’t find much literature on this issue. So you could look 

for literature that deals with charismatic leadership in other business sectors, and 

draw the relevant parallels to the hotel sector as well as acknowledging the probable 

differences. Fourth, again, there may be very little on the recruitment and selection 

of information technology professionals in China (your hypothetical chosen subject), 

but perhaps there is lots of material on the same subject in other geographical areas 

– the UK or the US for example. Once more, the relevant parallels can be drawn, and 

differences identified. And remember that, if there isn’t very much on your specific 

topic area, this suggests you have been successful in identifying something which is 

genuinely original! 

Some Tips on Electronic Searching 

Remembering the note of caution we attached to Internet searches in the Foundations 

of Management module book, here are three additional tips to assist you with electronic 

searches. First, using various ways to describe the topic helps to ensure your search. 

One example is that a search for material on downsizing might also include the use 

of search terms like ‘redundancy’ or even ‘derecruitment’. Second, remember that 

there is UK English and US English. Not only do UK English and American English 

spell words differently (e.g., the UK ‘behaviour’ and the US ‘behavior’), but they also 

use different words for the same thing (e.g., a UK chemist is a US drug store). If you 

don’t have any luck with one version of English in a particular search, try the other 

version. Third, a search facility might not recognise acronyms or abbreviations like 

ICI or TQM or BPR, depending on how it has been programmed. 

Finally remember that, for general enquiries concerning the University of 

Leicester Library and its services for distance learners, please complete the online 

enquiry form at 

http://www.le.ac.uk/li/distance/enquiry/reference_form.htm 

If you require help in finding subject information, you can also e-mail Andrew Dunn, 

our Management Information Librarian, direct at ad158@le.ac.uk 

OK, so let’s imagine you are now sitting down to read all of the books, journal 

articles, reports, conference proceedings, newspaper articles etc. that your extensive 

and comprehensive search has gathered. This can be a very daunting task indeed, so 

how should you go about it? 


Using the Literature 


As implied above, putting together a literature review entails you reading the relevant 

material and drafting a review fairly early on in the project process, but doing further 

searches to update your review and ensure it is as comprehensive as possible as 

you go along. Especially keep an eye out for new journal papers, books, newspaper 

articles etc., as we have said above. 

Taking Notes 

We also covered note-taking in Foundations of Management, of course. however, 

here is some extended and developed guidance on this issue, some of which repeats 

material presented there. 

When taking notes for your literature review, you need to read as widely as possible, 

but also to be selective. In other words, you may well gather material, particularly 

if you are trawling, that isn’t very relevant to your purposes. So before you start to 

read, look at the title, the back cover ‘blurb’ and/or the introduction if it is a book, 

or the abstract if it is a journal article. Ask yourself the following questions: 

• does this source relate to my research interests – is the context or content 

relevant to what I am researching? (if yes, then you should read it); 

• might it have been superseded by something more recent? (in which case try 

and find more up-to-date sources); 

• or is it a ‘classic’ or seminal text? (in which case you should read it); and 

• do other authors refer to it frequently? (in which case it is probably considered 

to be a classic text, or at least an important contribution to the area). 

Also note Robson’s (2002:52) observation that “My experience has been that quite 

often when key words have indicated a specific journal article, adjacent articles in 

the same journal have been of greater interest or relevance.” Further, with regard to 

books, use the contents pages and the index. Don’t just start at page 1 and keep 

reading until the end of the book – use the contents and index to ensure that you 

read the parts or chapters that are relevant to your project. 

When you have decided what to read and you are actually beginning the note-taking 

process, make sure you keep your research questions in mind. Keep asking yourself 

‘Can I use this material to explore my area of interest, and if so how?’ Record stuff 

which is relevant, not everything every author says. Even if you have chosen your 

texts carefully, not everything you read in every source is going to be pertinent to 

your area of interest. And write notes in your own words as far as possible. If you 

do want to use the author’s own words, then make sure you record the relevant 

section as a direct quotation using quotation marks, and take the page number 

down. Failure to do this could mean you being accused of plagiarism because chunks 



of other people’s words will probably end up without the appropriate attributions 

in your work. Plus remember that we are more interested in your words than other 

people’s as it is your words that prove you understand what you are talking about! 

Another important ‘to do’ when note-taking is to record all the relevant details for 

referencing purposes as you make your notes from each text – e.g., page numbers 

for direct quotations (as suggested above) and dates accessed for Internet sites. It’s 

also a good idea to compile a ‘running’ bibliography/list of references, adding to this 

as you read each text, to avoid a mad panic just before submission when you can’t 

find details for some of the texts that you have used! 

An additional top tip for note-taking is to formulate a working structure for your 

literature review as you read because 

“[t]rying to read everything then trying to write it up is a daunting task 

… it is unlikely you will appreciate the significance or possible location 

in the review of what you have read without some point of reference 

provided by a working structure.” (Gill and Johnson 2002:26) 

So think about how the text you are reading compares to others you have read. What 

are the main themes, concepts or claims that are emerging? What are the areas of 

agreement and disagreement between different authors in this regard? If you do 

this then you should be able to develop your notes into an effective literature review. 

It is really important to remember that the literature review is just that; a review. In 

other words, it should not be a list of authors or texts which reads ‘X says, Y says, Z 

says’. The idea is to present an overview of the literature which deals systematically 

with the main themes, concepts, claims and areas of agreement or disagreement 

between authors in this body of material. 

Below you can find an example of what we mean by this systematic overview, 

using the literature on whether men and women manage differently. A review of 

this literature would require you to have identified the two broad ‘camps’ in the 

available academic literature and therefore to organise your review in something like 

this format. 



Example 2.1: Sample (brief) literature review on gender and 

management styles 

There are two distinct camps in the gender and management styles 

literature, and I will review each in turn here. ‘Camp 1’ suggests that 

men and women manage differently because they are socialised 

differently. Rosener (1990) for example asked managers to describe 

their managerial style. her findings show that men said they adopted a 

‘transactional’ leadership style, based on the principle of exchange – in 

other words they gave their staff rewards or punishment for work done 

well or badly. These men also reported that they relied on positional 

authority in order to manage others – e.g., telling staff that they had to 

obey orders because they came from a manager. Women on the other 

hand reportedly used ‘transformational’ leadership – motivating staff 

through persuading them to commit to group/organisational goals, 

encouraging them to participate in decision-making, managing through 

personal qualities rather than by position, and trying to make staff feel 

good about themselves. Rosener also says these gender differences in 

management style are due to the differences in the ways boys and girls 

are brought up, and she concludes that transformational/‘feminine’ 

leadership is likely to be more successful in economically turbulent times 

than the transactional/’masculine’ style. In her 1997 book she extends 

this argument by suggesting that an important way of maintaining 

America’s corporate success and global competitiveness is to put 

women in senior positions in organisations, because their management 

style increases productivity, innovation and thereby profits through their 

aptitude for ambiguity and willingness to empower others. 

In summarising the available research on gender and management 

styles, Fagenson (1993) agrees with Rosener that the evidence indicates 

women prefer a transformational, interdependent style of leadership, 

instead of using status (like men). helgesen (1995) echoes Rosener 

and Fagenson – her research suggests that gendered management 

styles develop as a result of differential socialisation, and that women 

managers are consequently better at developing creativity, co-operation 

and intuition in others than men. Fagenson also emphasises women’s 

preference for managing via relationships as opposed to hierarchical 

position, and says that women listen and empathize much more 

than men. Again she asserts that feminine leadership ‘principles’ are 

becoming more influential because they suit today’s corporate realm 

better than male ‘warrior values’. 

‘Camp 2’ however suggests that, if men and women do manage 

differently, it is not just because of gender socialisation. Epstein (1991: 

151) points out that her own data from the legal profession and indeed 

her own experiences suggest women frequently display ‘combative’, 

‘punitive’ and ‘authoritarian’ (i.e., ‘masculine’) behaviour. She 



also notes ‘in-work’ variables like organisational size and culture as 

influencing management style, plus draws our attention to additional 

non-organisational variables in this regard like age, class and ethnic 

differences. Cohen (1991) says Rosener overlooks the fact that many 

of her female managers were responsible for professional staff who 

probably disliked a very directive, ‘masculine’ managerial approach. 

And if we overemphasise gender differences in management theorising, 

asks Gherardi (1995), how can we account for those men who prefer to 

manage in more democratic ways – like the 52% of male managers who 

said they preferred to use teamwork and a participative management 

style when surveyed by the British Institute of Management? White’s 

(1995) research into female executives also suggests these women were 

more different in their approach to leadership than they were similar. 

White attributes these differences to varying ages, experiences and 

expectations amongst this group. 

however, it is ‘camp 1’ that predominates. In other words, most writers 

on gender and management argue that there are gender differences 

in management style. This is despite the fact that “the majority of the 

academic empirical work supports the no-or-little-difference thesis” 

(Alvesson and Billing 1997:145 – emphasis added). 

So, having seen what the finished product might look like, let’s dwell a little further 

on writing up a literature review. 

Compiling a Literature Review 

When drafting and redrafting, again focus on your research questions and remember 

the literature review should be organised around themes, not presented as a list. In 

other words you need to avoid a review which looks like 

“the furniture sale catalogue, in which everything merits a one paragraph 

entry no matter how skilfully it has been conducted: Bloggs (1975) 

found this, Smith (1976) found that, Jones (1977) found the other, 

Bloggs, Smith and Jones (1978) found happiness in heaven.” (haywood 

and Wragg, cited in Bell 2005:100) 

Also make sure you cover all sides of the relevant arguments. In the example using 

the gender and management styles literature above, this review establishes that 

some authors believe there are gender differences in the ways that men and women 

manage, while others disagree, but that the ‘differences’ camp predominates despite 

the bulk of empirical evidence. 



In terms of structuring the finished version 2 , move from the general to the specific 

as follows. Begin with a short introduction laying out what the chapter will do – that 

is, give it a ‘top’ as discussed in Section 1. Then move to the main ideas, concepts 

and theories in the available literature – offer the kind of summary that would be 

found in the textbooks in the subject area. Summarise and contrast these key ideas 

then narrow down to more specific ideas; those most relevant to your research. 

here you will need to go beyond the textbook coverage, and use the research texts 

(monographs, edited books, journal articles etc.) to provide the necessary detail 

and complexity. highlight the areas where your research will contribute. That is, 

identify the weaknesses and gaps in the existing literature which your research will 

address empirically, or the new ways in which you will combine or evaluate literature/s. 

And finally offer a summary so as to provide a bridge to your methodology and data 

analysis chapters. So give your literature review a ‘tail’ – summarise the key points 

from the review and then suggest that these issues will form the basis for your own 

data gathering, as discussed in the methodology and data analysis chapters which 

follow. 

As the above suggests then, remember that: 

“you would also expect to return to the literature during the discussion 

and conclusion sections of your project report. You will want to present 

the significance of your empirical findings in the light of other people’s 

work and you will want to draw on other authors in arguing for the 

recommendations which you wish to make in the light of your findings.” 

(Jankowicz 2005:162) 

As suggested in the introduction and earlier in this section, you need to refer back 

to the literature review in your data analysis, so as to compare your work to other 

research as well as suggesting how the data answer your research questions. Do they 

confirm, amend or differ from the findings of other researchers? See Example 2.2 

below for an illustration of what we mean here. 

2 Also see Jankowicz (2005:164) for a useful example from an imaginary MBA 

project on leadership training in negotiation and sense-making skills. 



Example 2.2: An illustration of how literature is used in data analysis 

In halme’s (2000) research on ‘environmental learning’ in organisations, 

she conducted research at two Finnish firms – a packaging manufacturer 

and a paper maker. halme starts from the premise that the natural 

environment (e.g., resources like oil, gas, wood and water) has 

traditionally not been taken into account when strategic decisions 

have been made, because senior managers have tended to assume this 

environment is simply there to be exploited and that their primary focus 

should be on making profit. But this mindset is changing, so halme is 

interested in how senior managers start to include environmental values 

as a matter of course in their decision-making. 

halme is very careful to return to her earlier review of the relevant 

literature in her conclusions. She summarises this literature as arguing 

that “corporations must learn new ecocentric paradigms before they can 

be expected to produce environmentally sound performance” (halme 

2000:1087). her data on the other hand indicate that “cognitive-level 

environmental change does not inevitably precede behaviour change” 

(halme 2000:1087). So her Finnish managers learnt to be ‘environmentally 

friendly’ in their decisions by starting to take environmentally friendly 

decisions – i.e., they learnt by doing, as opposed to learning first and 

then doing. halme’s findings therefore differ from the orthodox wisdom 

in this area of management literature, and her data analysis makes this 

explicit. 

Finally, ULSM assessors want to see three main things in your literature review. 

First we are looking for evidence of reading: have you accessed and referred to 

the key texts, theories, concepts, models and findings? Is your review reasonably 

comprehensive and up to date? Second, have you understood what you have read? 

The aforementioned comparative, thematic approach is the key here, as is writing 

in your own words! Third, you need to evaluate the literature. how well do the 

various authors make their cases? What have they overlooked? What problematic 

assumptions do they make? Do they only use a narrow range of methods? Is their 

evidence convincing? Does this research tend to concern only a specific group of 

people or a specific geographical or organisational context? And how does all of this 

relate to your work? Evaluation thus leads into establishing your contribution as also 

discussed above. In summary, 

“What is required [from a literature review] is an insightful evaluation 

of what is known which leads naturally to a clarification of the gaps in 

the field and the way in which the proposed research is intended to fill 

them.” (Gill and Johnson 2002:26) 

NB you will also need to read the relevant literature to plan and write up your 

methodology, and make reference to this literature in the requisite places in your 



project – especially the methodology chapter, as also suggested in the introduction 

to this book. 

Summary 

i) Literature reviews are important in the Diploma project process for a variety 

of reasons. 

ii) There are a variety of different sources for subject-specific literature. 

iii) Your literature search will typically start wide and become more narrow as it 

proceeds. 

iv) When taking notes from the literature, keep your main purpose in mind. 

v) A literature review needs to be appropriately structured and to demonstrate 

reading, understanding and evaluation. 

vi) You should return to the literature you have reviewed in your data analysis to 

suggest how your findings relate to the existing body of knowledge in this 

regard, as well as how they answer your research question(/s). 


� Tasks 









References 




Alvesson, M. and Y.D. Billing (1997) Understanding Gender and Organizations London: 

Sage chapter 6 



Bell, J. (2005) Doing Your Research Project 4 th Edition. Buckingham: Open University 

Press chapter 6 




University Press chapter 4 


London: Sage chapter 3 

Fagenson, E.A. (1993) Diversity in management: introduction and the importance 

of women in management’ in E.A. Fagenson (ed.) Women in Management: Trends, 

Issues and Challenges in Managerial Diversity London: Sage pp. 3–15 

Foucault, M. (1986) The Foucault Reader P. Rabinow (ed.) harmondsworth: Penguin 

Gherardi, S. (1995) Gender, Symbolism and Organisational Cultures London: Sage 

Gill, J. and P. Johnson (2002) Research Methods for Managers 3 rd Edition. London: 


halme, M. (2002) ‘Corporate environmental paradigms in shift: learning during the 

course of action at UPM-Kymmene’ Journal of Management Studies 39(8):1087–1109 

Harvard Business Review (1991) ‘Debate: ways men and women lead’ January– 

February: 151–160. It incorporates Cohen, A.R. (p. 158); Epstein, C.F. (pp. 150–151) 

helgesen, S. (1995) The Female Advantage: Women’s Ways of Leadership New York: 

Currency/Doubleday 

Jankowicz, A.D. (2005) Business Research Projects 4 th Edition. London: Thomson 

Learning chapter 7 

Lechner, F.J and J. Boli (eds.) (2004) The Globalization Reader Malden, Massachusetts: 

Blackwell 


Practitioner-Researchers 2 nd Edition. Oxford: Blackwell pp. 45–65 

Rosener, J.B. (1990) ‘Ways women lead’ Harvard Business Review November– 

December: 119–125 

Rosener, J.B. (1997) America’s Competitive Secret : Women Managers Oxford: Oxford 

University Press 




5 th Edition. harlow: Financial Times Prentice hall chapter 3 

White, J. (1995) Leading in their own ways: women chief executives in local 

government’ in C. Itzin and J. Newman (eds.) Gender, Culture and Organizational 

Change: Putting Theory into Practice London: Routledge pp. 193–210 




section 3 

Formulating Research Questions




SECTION 3 

Formulating Research Questions 


having studied this section of the module you will be able to: 

• recognise the difference between your research topic and your 

research question(s) 

• understand how personal interest should affect your decision(s) 

• recognise the benefit of keeping an open mind about where your 

research might take you 

• understand the value of flexibility and feasibility, and how they 

will help you maintain focus. 

Introduction 

It is crucially important that you select a well focused and thought out research 

question(/s), and that you have a clear rationale for why this question(/s) is important. 

Indeed one of the major deficiencies we see in students’ research projects is a lack of 

clarity as to what is being researched and why this question(/s) is significant. So let’s 

delve a little further into the challenges of formulating effective research questions. 

Research Topics versus Research Questions 

Your research topic or area is not the same thing as your research questions. The 

topic might be (here goes with that list again!) organisational structure, organisational 

culture, relationship marketing, charismatic leadership, stress, productivity, employee 

motivation, the supply chain, auditing … or whatever. Your research questions, 

however, indicate gaps or uncertainties in our knowledge about a specific topic 

– they should aim to discover things that we do not know already. Remember 

that, as we established in Section 1, “[t]o ‘research’ is to seek answers that involve 

understanding and explanation” (Williams and May 1996:7 – emphasis added). 

To give a concrete example, your topic might be work-life balance. The sector you are 

interested in might be the British financial services industry, and you are also keen to 

research managers as opposed to other groups of workers. Your research question, 



which indicates what you want to find out about this topic, might therefore be 

posed as follows: 

• To what extent is work-life balance a problem issue for British male managers 

in the financial services industry as compared to British female managers? 

This is interesting because the vast majority of the relevant literature discusses worklife 

balance as a problem that is experienced mainly by women – due to the fact 

that in heterosexual partnerships women still retain most of the responsibility for 

domestic labour and childcare, even if they work full-time. 

Sub-questions, which are advisable because they narrow the focus of the project 

even more, could then become: 

• Do British male managers in this industry report experiencing difficulties in 

achieving work-life balance? If so, what are the causes? And what might the 

solutions be? 

• To what extent do British male managers in this industry see work-life balance 

as a significant source of stress? If so, what are the implications for their 

health, quality of life and performance at work? 

Please note that the more you can narrow the research questions down – to country 

of focus, respondent group, sector/industry/organisation etc. – the better and the 

easier your research will be. We return to this issue below. 

Now let’s move on to the things you need to consider in choosing a topic, and 

putting research questions together. 

What To Think About in Topic and Question Selection 

Personal Interest 

The first and most vital issue here is to make sure the topic and question(/s) are 

of personal interest. You need to select a focus which will hold your interest 

throughout the duration of the project process. We also drew your attention to this 

in Section 1. So you might choose a topic/questions based on your own (work/ 

non-work) experience – e.g., Fantasia wrote about worker solidarity in the US as 

a result of working at a small iron foundry when a wildcat strike occurred (cited in 

Robson 2002:49). One of the authors of this module book has studied the body in 

organisations for an extended period of time because she is extremely body conscious 

herself! Or perhaps you have been particularly attracted to an issue covered in one 

of your Diploma modules, or there is a current news story in your country to do 

with an aspect of management which has caught your eye. Alternatively, you may 

wish to do something influenced by your political or moral beliefs – e.g., a study of 

environmental management if you are a member of Greenpeace. 



Keeping an Open Mind about Findings, Methods and Data Analysis 

Although as we have said above it is crucial that you choose a topic and question/s 

that interest you, so you can keep your interest and motivation going, it is also very 

important that you do not pre-judge your findings, or allow your interest or expertise 

in a particular method or approach to analysing data to influence the research question 

that you select. First you need to keep an open mind about what the answer to your 

research question/s might be, because otherwise your preconceptions will mean you 

are likely to ignore important data or to misinterpret the data that you do gather. 

Second, research methods and data analysis techniques should be chosen to fit the 

research topic, not the other way around. 

Using the Subject-Specific Literature 

Once you have decided on a topic area, it is a good idea to look at the subjectspecific 

literature for suggestions for future research. As Section 2 establishes, this 

material may actually outline questions which still need to be answered about the 

relevant topic. A critical review of the literature also allows you to spot gaps and 

weaknesses. We have mentioned the idea of making a contribution several times 

already in this module book, and a discerning scan of the literature allows you to see 

where you might be able to add something new by looking at something that hasn’t 

been studied before, by using an unusual method, by gathering data from a different 

kind of sample or by bringing together and/or evaluating literature in a novel way. 

It also allows you to ensure your research question is of contemporary relevance – 

something in which today’s management academics and managers themselves will 

be interested (as already discussed in Section 1). 

Ensuring ‘Real World Value’ 

Again we have alluded to this issue in Section 1. ‘Real world value’ is Robson’s 

(2002:56) term – he adapts it from Campbell et al. They refer to successful management 

research having relevance or usefulness to managers, to its potential to address 

specific organisational problems or improve organisational processes or procedures. 

Not only does this mean that the research has a use beyond adding to academic 

knowledge but it might also give something back to the organisation/s where you are 

doing the research, and/or which is sponsoring you. It might also make access easier 

if you can promise some kind of report which will make specific recommendations 

for organisational improvements (more on this in Section 6). But do remember that 

you are not acting as a consultant or aiming to produce a management report: your 

project must address issues of academic significance and be written according to 

academic protocol, as (once more) we established in Section 1. 

Brainstorm It 

Again this is something that we discussed in Foundations of Management, but 

brainstorming can also be used to come up with research questions. Gill and Johnson 

(2002:19) give a useful example around the topic of managerial stress. Remember 

you can brainstorm with others or alone. The process is to write down any questions 



as they come to mind, however silly they may seem, under a broad topic heading. 

Only when you have run out of ideas should you start to evaluate them in terms of 

how effective they are likely to prove as research questions. 

Focus and Feasibility 

Focus and feasibility are also really important. Students often make the mistake of 

choosing research questions which are much too broad. This means that you will not 

have the time or the necessary resources in terms of money or equipment to actually 

find out the answers to these questions. You need to keep your questions as narrow 

as possible – think about the range of literature you will have to consult and the 

variety of data you will need to collect - do you have time? Do you have the necessary 

money and equipment? Are the literature and data you need likely to be available to 

you? Can you do analytical justice to them even if they are? 3 

Gill and Johnson (2002:16) give the example of a project seeking to compare the role 

of human Resource departments in the UK and the US. But, say you were a UK-based 

student, how would you get the US data? If you can’t use questionnaires, you don’t 

have the money to get to the US to do interviews and the data aren‘t available in 

secondary form, then how would you do this research? Another example might be 

suggesting that you want to evaluate hRM in a particular organisation or compare 

two organisations in terms of their approach to hRM. But hRM encompasses many 

different activities and responsibilities, so it would be much more sensible to focus 

on one of its aspects, such as recruitment or employee involvement or performance 

management. Similarly, some students have said in the past that they want to look 

at, say, TQM in ‘the Chinese economy’ or ‘the Greek economy’. however, this again 

is much too broad – we would suggest choosing a particular sector like banking or 

hotels or car manufacturing, or even better one organisation. however, you still need 

to work towards a balance here – focus and feasibility are crucial, but as Campbell 

et al. (cited in Robson 2002:56) also point out, the cheapest, quickest and easiest 

research isn’t always the best research! 

Example 3.1 offers an example of how a research question that might look broadly 

OK actually wouldn’t work very well in practice. 

3 Please also remember that your project must be entirely self-funded: in other 

words, the School of Management cannot provide any money or equipment to assist 

you in your research. 



Example 3.1: Assessing a research question for ‘researchability’ 

Imagine you wanted to ask the following research question: ‘how well 

has the British car industry performed compared to the German car 

industry?’ As it stands, although it might look all right at first glance, 

this question would be difficult if not impossible to research sensibly 

and systematically. It needs substantial refining and revision. 

For example, what is meant by performance here? Global market 

share? Return on investment? Customer satisfaction? Error levels 

in manufacturing plants? Quality of working life for employees? 

Environmental ‘friendliness’ (etc.)? You would need to specify this, as 

your project could not hope to cover all of these performance indicators. 

Also, what time period does the research cover? Again this would need 

to be set out. Post-German unification perhaps? In fact this is a very 

long period of time – some twenty years or so at the time of writing – 

so the sensible approach would be to narrow it even further, perhaps 

to 2003–2008. 

Finally, how is ‘British’ or ‘German’ car industry defined here? Nissan, 

for example, has some of its operations in Sunderland in the North 

of England, but it is a Japanese-owned car manufacturer. Similarly, 

Mercedes-Benz has a plant in Tuscaloosa, Alabama in the US but it is 

a German-owned car manufacturer. So, once again, you would need 

to specify whether you simply meant German-owned/British-owned car 

manufacturers, or some other way of defining car industry in this regard 

(the narrower the better!). 

Enhancing Career-Relevant Knowledge 

We established in Section 1 that one of the reasons why we ask you to do a PMP is 

to further your knowledge so you can use it in improving your work-related practice 

and progressing your career. So choosing a topic/question/s that will be especially 

relevant to you in this regard is a good idea. For example, a former student of ours 

chose to research the extent of and reasons for tax evasion in her country of origin 

as her project topic because she wanted to get a job in the relevant Tax Inspectorate 

when she graduated. 

Symmetry of Potential Outcomes 

What this means in simple terms is that you need to make sure that your findings will 

be interesting whatever the outcomes of your research. For example, say your key 

research question was ‘Does the increasing use of e-mail represent a stress factor in 

the Ghanaian banking sector?’ If your findings suggested that the increasing use of 

e-mail was not experienced as stressful by Ghanaian bank workers, then this would 



not be particularly interesting. This question therefore does not have symmetry of 

potential outcomes. But by rephrasing the question as ‘Does the increasing use of 

e-mail represent a stress factor, a benefit or both in the Ghanaian banking sector?’, 

then whatever the findings showed would be interesting. 

How Sensitive is Your Topic/Question/s? 

This can present a possible problem in formulating topics and questions. Easterby- 

Smith et al. are quite right to say that “Access to companies can be obstructed by 

managers if they consider a piece of research to be harmful to their, or their company’s, 

interests” (2008:2). We will talk more about getting access to organisations for 

empirical research in Section 6, but this might represent a real challenge if you 

choose a topic or question/s that seems to be sensitive. For example, the question 

of the effectiveness of drug testing at work could be seen as very sensitive on a 

personal, individual basis because it relates to confidential medical information and 

also to employees’ private, leisure time behaviour. On the other hand, a project 

could be regarded as commercially sensitive – e.g., something which focused on 

branding, unique selling points and/or competitive strategies. There are ethical issues 

here too relating to not invading respondents’ privacy or putting them under any 

duress. Also more on this in Section 6, but bear in mind Gill and Johnson’s (2002:16) 

citation of a researcher who told them that he had a lot of difficulty getting access to 

managers to talk to them about their stress levels, and he felt this was because they 

were so stressed! 

How About You? 

Equally, consider your own capabilities. Doing a PMP is intended to develop your 

skills and abilities in a range of different ways, as discussed in Section 1. however, it’s 

still unwise to choose something which is a long way away from your ‘comfort zone’ 

– e.g., don’t choose a research question which would require a lot of quantitative, 

statistical work if you are not very numerate. 

Research Questions Can Change 

And, finally, don’t forget that research questions are often iterative. In other words 

your specific research question(/s) and the approach that you choose to answer it 

may well alter as your project proceeds – for example, personnel at the relevant 

organisation might change, requiring access renegotiation, or you could discover 

new angles or concepts on the issues as you do more reading. This is not a problem, 

but we strongly recommend that you check out any changes you plan to make with 

the tutor on the relevant Blackboard project support forum. 

Finally, and just to repeat ourselves one more time, a clearly defined research 

question(/s) is also much more likely to generate successful research, because your 

pathway through the research process will be easier to identify as a result. In other 

words, your research question should help you to decide what you need to read and 

what kind of data you need to gather, where this data gathering will happen, when 

it will happen and how it will happen. Also remember that “[a]ll enquiry involves 



drudgery and frustration and you need to have a strong interest in the topic to keep 

you going through the bad times” (Robson 2002:49). In other words, if you start off 

not being interested in your topic then this is a sure-fire way to produce a lacklustre, 

unsatisfactory project! 

Summary 

i) Research topics are the broad areas that we are interested in examining. 

ii) Research questions represent gaps in our knowledge about those topics. 

iii) Research sub-questions are a useful way of narrowing down the focus of your 

research. 

iv) The most important issue here is your interest in the topic/ question(/s). 

v) The second is probably focus and feasibility … but there are many others! 


� Tasks 









References 







University Press pp. 82–89 



Easterby-Smith, M., R. Thorpe and A. Lowe (2008) Management Research 3 rd Edition. 

London: Sage chapter 1 

Gill, J. and P. Johnson (2002) Research Methods for Managers 3 rd Edition. London: 



Learning chapter 2 


Practitioner-Researchers 2 nd Edition. Oxford: Blackwell pp. 45–65 



Williams, M. and T. May (1996) Introduction to the Philosophy of Social Research 

London: UCL Press chapter 1 




section 4 

Quantitative and Qualitative 

Methods – A Selective Review




SECTION 4 

Quantitative and Qualitative Research 

Methods – A Selective Review 



• distinguish between quantitative and qualitative methods 

• outline some of the uses of structured interviews, self-administered 

questionnaires, structured observation, unstructured observation, 

semi-structured and unstructured interviews 

• choose the method(s) that is right for you and your project. 

Introduction 

As suggested in the introduction to this module, there are multiple methods in the 

social sciences and we can’t cover all of them in a module book like this. Instead 

we will explore issues related to the ones which, based on ULSM experience, most 

students of management use. This section therefore deals with examples of each 

method in management research and with choosing a method. 

Distinguishing Between Quantitative and 

Qualitative Methods 

Let’s begin with a quick distinction between quantitative methods and qualitative 

methods. Quantitative methods tend to generate data expressed numerically, 

which are analysed statistically. Quantitative data, then, are either gathered as 

numbers (e.g., age, income, number of children) or turned into numbers after they 

have been gathered (e.g., by coding responses to closed questions 4 or content 

analysis 5 of qualitative data). Analysis of these data is usually conducted through 

statistical operations, so one is looking for the relative frequency of particular 

variables, for what is happening on average in these data. 

4 A closed question is one that has a fixed set of responses such as yes/no/don’t 

know. More on this issue in Section 5. 

5 More about content analysis in Section 8. 


See Section 7 for more on analysing quantitative data. 


Qualitative methods, on the other hand, tend to generate data expressed in words, 

which are analysed conceptually. here data are collected as words – e.g., via open 

questions 6 – and analysis is conducted through classifying these data into categories or 

themes and then trying to understand these themes via theoretical conceptualisation. 

Qualitative data is often understood as providing a ‘richer’ description of the relevant 

empirical site. Again, more on analysis of qualitative data in Section 8. Now let’s move 

to examine two examples of quantitative methods in more detail. 

Quantitative Methods 

The two methods we focus on here are structured interviews and self-administered 

questionnaires. These are also known as ‘survey’ methods, because they involve 

collection of standardised information, usually from a large, representative sample 

(more on sampling in Section 5). 

The researcher who chooses these methods typically wants to be objective, and to be 

detached from their respondents. Indeed with self-administered questionnaires the 

researcher may never meet the people who take part in their research. Both methods 

consist of asking the same questions in the same order to every respondent. The 

difference is that with the structured interview the researcher asks the questions 

verbally, either face to face or over the phone, and records the respondent’s answers 

in writing or electronically. With SAQs the respondent reads the questions and fills 

the answers out themselves by writing or typing them on to a pre-prepared form. 

These methods tend to rely on closed questions so responses can be quantified (i.e., 

counted). 

What follows is an example of the use of each method in management research. 

Structured Interviews: An Example from Management Research 

You may also see these referred to as formalised, standardised or respondent 

interviews. An example is Oliver et al. (2002) on the relationship between psychological/ 

individual factors, organisational variables and occupational accidents. The data 

were gathered in the Valencia region of Spain. Some 525 structured interviews 

were conducted with a random sample of workers attending for annual medical 

checks at the Valencia health and Safety Executive. The literature on occupational 

accidents argues that they are caused by organisational variables and psychological 

or individual variables. Oliver et al. look at how these two types of factor interrelate 

in this regard. They set out to test a hypothetical 7 model of such interrelations which 

they developed based on the relevant literature. 

6 An open question allows any answer that the respondent – the person 

answering the question – sees fit to give. Again more on this in Section 5. 

7 We discuss hypotheses further below. 



The interviews asked respondents questions about 

• their biographical data like their age and their job; 

• their experience of different types of occupational accidents over the preceding 

two years; 

• organisational variables like their perceptions of how supportive colleagues 

and supervisors were and how seriously management took the issue of safety; 

• the quality of work conditions like levels of noise, lighting etc.; 

• their general health; 

• and their own safety behaviour – e.g., did they ever take shortcuts when 

using equipment? Did they always wear the correct safety gear? 

Oliver et al.’s findings suggest that organisational variables are a key issue in terms of 

how workers assess workplace hazards – so safety behaviours are not influenced by 

the riskiness of the work as such, but more by the safety procedures in place. human 

factors are identified here as significant as opposed to ‘hardware’ or technical factors 

– and the conclusion is that there is a real need to actively manage safety and to stamp 

out unsafe practices and shortcuts. Levels of stress were also shown to mediate – i.e., 

either reduce or enhance – safety behaviours, so social support is also an important 

issue in reducing accidents at work according to these findings. 

Self-Administered Questionnaires: An Example from Management Research 

Naudé et al. (1997) used self-administered questionnaires to examine the extent to 

which managers use a range of statistical techniques in their jobs. The authors suggest 

that these techniques allow more reasoned and rational decisions to be made by 

managers, and say this is supported by the existing research. But how frequently are 

these techniques actually used in management decision making? The available data 

suggest they are not necessarily used very frequently, despite rising numbers of MBA 

graduate numbers and the central place that quantitative techniques have on any 

MBA programme. So Naudé et al. wanted to measure how much these techniques are 

used by MBA graduates at work – and they selected 13 commonly taught techniques 

including forecasting models, regression and correlation, and probability analysis. 

The researchers sent an SAQ to 3419 respondents in the UK, South Africa and New 

Zealand. A total of 1219 were returned, giving a response rate of 36% 8 . The SAQ 

asked 

8 As should be obvious, response rate is the numbers of people responding as 

against the total number of SAQs sent out. here then 3419 SAQs were sent out, and 

1219 people responded. The calculation thus becomes 1219 divided by 3419, which 

gives a response rate of 35.65% (rounded up to 36%). 


• for biographical data like the respondent’s management level; 


• which quantitative techniques they used at work and why (or why not, where 

techniques weren’t used); 

• which they used most often; 

• whether the respondent had any suggestions regarding the appropriate 

content of an MBA quantitative methods module. 

The results show consistent usage of a range of methods at between 60–70% in 

each country and indicate that respondents had broadly the same reasons for using 

and not using various methods. There was also a relatively low level of awareness of 

some techniques in particular – e.g., non-parametric tests and multivariate models. 

Naudé et al. suggest this might either be because the respondents had forgotten 

these methods from their MBA studies, or they had never been introduced to them. 

New Zealand respondents tended to use all methods more frequently. Again Naudé 

et al. suggest this might be because quantitative methods are taught differently on 

MBA programmes in this country, or because the NZ respondents were younger on 

average than the other two cohorts, so may have needed more quantitative techniques 

because of their more junior (and so more operational) roles. 

Regarding recommendations for MBA quantitative methods modules, Naudé et al.’s 

experience is that these are unpopular with students. however, their data suggest 

none of the techniques should be excluded, although some may be better delivered 

as part of an elective. This suggests an existing overlap between supply (what MBA 

tutors think is needed) and demand (what MBA students want). Other research they 

refer to argues that the practical application and relevance of these techniques to 

real-life management work may be neglected on MBA modules. Naudé et al. say more 

research is therefore needed on how to teach statistics to postgraduate students of 

management. 

NB Quantitative research methods also include the use of quantitative secondary 

data, i.e., existing data, usually from large and representative samples, and expressed 

in numerical and/or statistical form such as the UK Labour Force Survey or Census, 

performance data from company annual reports, organisations’ absence records or 

quantitative data from existing academic publications. 

Quantitative/Qualitative Methods 

All methods can be varied to become more or less quantitative or qualitative where 

necessary, more of which below when we discuss hybrid methods. however, 

observation, which involves watching people go about their everyday business, is 

one method that is routinely either quantitative or qualitative, depending on how it 

is used. Structured observation is a quantitative method, whereas non-structured 

observation is typically qualitative. 



Structured observation is usually intended to be objective, detached and aims to 

quantify behaviour. So the researcher uses a ‘tick box’ observation schedule where 

they record how many times people do something, the order in which they do it and/ 

or how long it takes to do it. Non-structured observation wants instead to understand 

naturally occurring activity with the aim of understanding it in ‘the round’, in all 

its complexity, as opposed to the attempt to focus on specific behaviours and to 

quantify them as in structured observation. 

But observation can also be classified as either participant (also known as 

ethnography) – i.e., actually living or working amongst the people in whom you 

are interested and taking a full part in their daily activities – or non-participant 

(also called passive) observation. In the latter the researcher simply watches naturally 

occurring activity without taking part. It would be difficult to conduct participant 

observation which was also structured, however, and this is usually not the intention: 

instead the researcher wants to immerse themselves in others’ experiences. 

Still another way in which observation can be classified is as covert (where respondents 

do not know they are being watched, because the researcher is pretending to be 

someone else, like a new colleague, or because respondents are being recorded 

via the use of technology like CCTV) versus overt (where they do know they are 

being observed). Both covert and overt observation can be either structured or nonstructured 

and either participant or passive. 

Thinking point 4.1: Covert observation 

how would you feel if you realised that someone you thought was a 

new recruit to your organisation had in fact been observing you and 

your colleagues covertly for the last six months, and was going to write 

up the resulting data for publication in an academic journal? 

We will return to the ethics of covert observation in Section 6. 

What follows is one example of structured observation, and two examples of nonstructured 

observation, in management research. 

Structured Observation: An Example from Management Research 

Martinko and Gardner (1990) take henry Mintzberg’s well known non-participant 

observation study of five CEOs, published in the early 1970s, as their starting point. 

This previous study argues that management work is “varied, brief, fragmented and 

highly interpersonal” (Martinko and Gardner 1990:329). But the literature suggests 

there is still some controversy about Mintzberg’s claim in terms of the extent to 

which management work is non-systematic and thus hard to control or plan. 

Martinko and Gardner replicate 9 Mintzberg’s study using structured observation. 

9 That is to say, they repeated his methodology on a different sample of people. 



They also aim to explore relationships between management behaviour, management 

effectiveness and environmental and demographic variables, since they think the 

last two issues have been neglected in the relevant literature. Martinko and Gardner 

start with three hypotheses based on this literature: 1. that management work is 

varied, brief, fragmented and highly interpersonal; 2. that there are differences in 

behaviour between effective and less effective managers (i.e., in terms of what they 

do and how they do it); and 3. that biographical and environmental differences affect 

management behaviour. 

Using the categories of management activity developed by Mintzberg, Martinko and 

Gardner studied 41 US school principals, of which 22 were high and 19 moderate 

performers according to data like their students’ performance in national tests. 

Environmental and demographic variables included the grade level covered by the 

school (i.e., elementary or secondary), staff numbers in the school and levels of 

urbanisation in the surrounding area. Minute by minute descriptive observation was 

performed but separate behavioural events were also classified according to how 

long they lasted, who initiated them (e.g., the principal themselves or someone else) 

and what kind of event they represented (e.g., scheduled meetings, unscheduled 

meetings, phone calls etc.). The average length of observation was 6.7 days per 

principal. All observation was also non-participant and overt. 

Data were then coded according to the purposes of the event – e.g., scheduling, 

receiving information, giving information – and Mintzberg’s managerial roles – e.g., 

was the principal acting as a figurehead, a disseminator, a negotiator etc. during 

this particular event? The results confirm that management work is varied, brief, 

fragmented and interpersonal – i.e., much of it is spontaneous, non-scheduled and/or 

not initiated by the manager themselves. Moreover, 50% of these managers’ time was 

spent on interpersonal and informal communication. So hypothesis 1 was supported. 

Managers’ behaviour does not however seem to be linked to performance in these 

data (so hypothesis 2 is not supported), although environmental and demographic 

factors were linked to different management behaviours (supporting hypothesis 3). 

Unstructured Observation: Two Examples from Management Research 

The first example we use here is also a classic piece of participant observation. Roy 

(1960) worked as an assembly line operative (i.e., doing very monotonous, unskilled 

work) with three other men – Ike, George and Sammy – for an extended period of 

time. Ike, George and Sammy thought he was a university student who needed extra 

money; so this observation was also covert. As Roy (1960:205–206) suggests, “My 

account of how one group of machine operators kept from ‘going nuts’ in a situation 

of monotonous work activity attempts to lay bare the tissues of interaction which 

made up the content of their adjustment.” In other words, his focus was on how this 

group of men actually coped with such boring, repetitive work day after day. 

To be more specific, after a while, Roy began to notice that the three men had 

developed various ‘times’ which helped them to get through the shifts. These ‘times’ 

were frequent but short. They marked the actual passage of the working day because 

they always took place at the same time in the shift. But more importantly they 



functioned to provide a focus of interest and discussion until the next ‘time’ took 

place – even though these times were the same every day! For example, at Banana 

Time (also the title of the article), Ike would steal and eat the banana that Sammy 

always brought for his lunch. Sammy would complain, and George would mildly 

remonstrate with both of them. 

There were also ‘themes’, which were standard patterns of interaction in which the 

group engaged. These were not, however, as predictable in their occurrence and 

regularity as the ‘times’. ‘Themes’ included the ‘professor theme’. George, who was 

both the formal leader of the group and highly respected by his colleagues, had a 

daughter who had married the son of a college professor. George would regularly 

regale the group with stories of the wedding and his Sunday walks with the professor, 

or dinner with the professor’s family. Roy felt the respect shown to George by Ike and 

Sammy was at least in part due to the ‘professor theme’. 

Our second example focuses on shopfloor humour and took place in the components 

division of a lorry-making factory in North West England. Two hundred and fifty men 

worked in the division. Collinson (1988) explores how the men’s joking reflected and 

reinforced their values as well as representing resistance to management, a way to 

control each other and so on. Examples of the humour included: 

• nicknames – e.g., ‘Electric Lips’ couldn’t keep secrets, and ‘Silver Sleeve’ didn’t 

use a handkerchief (Collinson 1988:185, 195) 

• pranks such as an amateur weightlifter being challenged to lift Allan, a colleague 

who could apparently make himself heavier “at will”. The weightlifter failed 

because someone else had nailed his shoes to the board he was standing on, 

so he was “trying to lift Allan, the board and himself” (Collinson 1988:188); 

• initiation rites for new workers like being sent to another colleague to ask for 

“a long stand”, and being asked after standing waiting for some time “Is that 

long enough?” (Collinson 1988:189); 

• jokes at each other’s expense – e.g., one colleague sending a rate-fixer to study 

another rate fixer’s job, “with the implication that he was lazy” (Collinson 

1988:195); 

• jokes at the management’s expense, like a story about three foremen not 

being informed about a training course in communication skills (Collinson 

1988:186). 

The humour, as is obvious from the examples above, was very unforgiving and harsh. 

It was also very masculine, indeed macho, sexual and contained lots of epithets. 

In a similar vein to Roy, Collinson emphasises that these men did repetitive and 

mundane work, worked the longest hours in the factory and had very poor terms 

and conditions – all of which suggested they “were the least valued and most easily 

disposable of employees” (Collinson 1988:185). So they had developed a culture of 

humour which preserved their self-esteem through, for example, emphasising their 



masculinity compared to what they called the “twats and nancy boys” (Collinson 

1988:186) in the offices. These office workers had (as the men saw it) no freedom to 

joke around. 

The humour, in addition, helped with boredom, and was a means to achieve acceptance 

from colleagues. If an employee could both ‘dish it out’ and ‘take it’ he became ‘one 

of the gang’. If someone didn’t join in then they were ostracised. Finally humour 

was used as a disciplinary mechanism to bring colleagues into line – e.g., someone 

would often become the butt of a cruel joke if they were seen not to be working hard 

enough, because part of the men’s wages was based on a collective bonus system. 

We should also note that Collinson simply watched day to day interaction in this 

environment – i.e., this observation was non-participant – and that the men were 

aware that he was an academic researcher, so the observation was overt. 

Qualitative Methods 

The two methods we discuss here are semi-structured and unstructured interviews. 

You may also see these referred to as informal, non-standardised, informant, nondirective, 

open-ended or in-depth interviews. In some ways they might be seen as 

variants of each other; the difference simply being the level of structure employed in 

the interview schedule. 

These methods are (usually) face to face interactions between researcher and 

respondent during which the researcher wants to cover particular topics and ask 

specific questions, but (unlike structured interviews) where the order in which the 

questions are asked and the wording used depends on the respondent. For example, 

the respondent might highlight issues in answering one question which aren’t due to 

be covered until later in the schedule, but here the researcher would move straight 

to asking about these issues as this follows the respondent’s logic. So the interview 

is effectively led by the respondent as opposed to being led by the researcher, as 

in the structured interview. Open questions are typically used and there is usually no 

particular emphasis by the researcher on objectivity or detachment – instead they aim 

to build up a rapport with their respondents, to really get to know them during the 

interview process. 

Such interviews (along with non-structured observation) are also known as ‘case 

study’ methods, because they gather rich and detailed data, usually from a small 

sample which tends not to be representative of a wider population. What follows is 

an example of the use of each method in management research. 

Semi-Structured Interviews: An Example from Management Research 

Our first example is Brewis (2000), who discusses the results of semi-structured 

interviews with 16 women about their experiences of their bodies at work. Because 

the sample was small and a combination of convenience, purposive and self-selection 

sampling techniques was used (more of which in Section 5), Brewis does not identify 



it as in any way representative of working women more generally. Instead the data 

gathered are treated as a series of “impressions and interpretations of the experience 

of having a female body” (Brewis 2000:171). These data suggest that the respondents’ 

body images are produced by exposure to others’ reactions, others’ bodies and 

cultural images of women’s bodies (e.g., in advertisements). So these women have 

‘learnt’ about their bodies, about how attractive and socially acceptable they are, by 

being in the world, seeing other people’s bodies and noting how other people react 

to their bodies. 

The respondents also mainly felt that their bodies didn’t ‘measure up’ to cultural 

expectations about women’s bodies. Nonetheless, they didn’t always spend much 

time rectifying this. For example, the mothers in the sample said they spent much 

more time on preparing their children’s bodies for school than in preparing their own 

bodies for work, because this is a crucial part of being a mother. Other women who 

spent very little time getting ready in the mornings seemed to regard bodily titivation 

and preparation as ‘feminine’ activities, whereas they needed to be seen as more 

‘masculine’ (objective, rational, logical, assertive etc.) in their organisational roles. 

however these respondents still thought they were labelled by their bodily sex at 

work. For example, two women who defined themselves as overweight believed 

colleagues saw them as lacking discipline or self-control, because women almost 

have to ‘embody’ these characteristics in the workplace whereas an overweight man 

would not be judged in this way. Some respondents worked to ‘blend in’ as a result 

– e.g., wearing masculine suits to play down their biological sex. Others wanted to 

stand out a bit more and used their bodies to their advantage. So several respondents 

said that female bodies, being smaller than men’s in the main, can be beneficial in 

terms of managing conflict situations at work (such as an angry customer) because 

people see them as less threatening. 

Unstructured Interviews: An Example from Management Research 

Cashmore (2002) reports data from unstructured interviews, some undertaken oneon-one, 

others in the form of focus groups where several respondents discuss the 

relevant issues as a group as prompted and guided by the researcher. his research 

topic is ethnic minority officers’ views of cultural diversity in the police service. One 

hundred African Caribbean and South Asian police officers in three English regions 

– the West Midlands, Norfolk and Derbyshire – took part. Each area is very different 

from the others in terms of ethnic minority population. The research took place 

18 months after the February 1999 publication of the Macpherson Report into the 

racist murder of black teenager Stephen Lawrence in London in 1993. This report 

suggested that the Metropolitan Police are institutionally, systematically racist, which 

in part accounted for their failure to bring anyone successfully to trial for Stephen’s 

murder. It made two key recommendations: that the police should seek to recruit 

more ethnic minority officers and that better cultural diversity training was needed 

in the police service. The report also put targets in place regarding the recruitment of 

ethnic minority officers. 



Cashmore focuses on how ethnic minority officers themselves interpret the Macpherson 

recommendations. Interestingly, their reaction was mainly negative. First, officers 

suggested that initiatives like these are often more about public relations and the 

external image of the police service than actually addressing institutional racism. 

They also pointed out that such initiatives have tended to fail in the past. So what 

looks like action or progress in fact is not. Second, they were not sure the targets set 

were realistic. Third, officers commented that targeting ethnic minority people for 

recruitment purposes might be tokenistic, so that new recruits who come into the 

service as a result of these initiatives might not be as motivated or as able as existing 

officers. 

Fourth, they suggested the police service culture tends to dilute any existing ethnic 

minority affiliation held by its members anyway. Indeed these officers felt they now 

saw the world through ‘white eyes’ because they had been working in the service 

for some time. Fifth, they commented that diversity training in the service focuses 

on dealing with the public, but needs to focus more on how police colleagues relate 

to each other … and so on. Cashmore makes several recommendations to the police 

service on the basis of these data, including the fast tracking of ethnic minority officers 

to more senior positions, to demonstrate that they can ‘make it’ both to potential 

ethnic minority recruits and the British media. however, he does acknowledge that 

this is likely to generate resentment from white officers. 

NB Qualitative research methods also include the use of qualitative secondary 

data, i.e., existing data, in descriptive/narrative form, such as minutes of meetings, 

text from annual reports, newspaper articles, court reports, qualitative data from 

academic publications etc. 

hopefully the section so far has given you an idea of what these various research 

methods can be used to do, which should go some way towards helping you to 

choose one. The next section deals with the process of choosing in much more depth. 

Choosing a Method 

One key issue in choosing a method is what its strengths and weaknesses are, which 

we deal with in a task for this section (see the Blackboard module site), and how 

these strengths and weaknesses relate to your own research. here though we cover 

some other considerations, as follows. 

What Kind of Research Are You Interested In Doing? What Kind of Stance Do You 

Want to Take? 

First, are you more interested in quantities or qualities? An interest in quantities or 

frequencies of social phenomena – i.e., how many there are, how widespread they 

are, where they are, when they occur, the extent to which they occur etc. – imply 

that you want to count or measure these phenomena. In other words, you want to 

quantify them. 



For example, perhaps you want to 

• count different types of organisational structure or organisational culture in 

a particular organisation or industry 

• quantify the extent to which a particular organisation engages in relationship 

marketing 

• measure how much charismatic leadership particular leaders display 

• quantify employee stress or motivation or productivity in a given organisation 

• measure the efficiency of a specific market 

• quantify how different types of auditor’s reports affect share prices in a 

particular industry 

• measure the ways in which a particular supply chain adds value to a good or 

service … etc. 

If you are interested in quantities then a quantitative, survey method would work best 

for you, like a self-administered questionnaire, a structured interview, quantitative 

secondary data or structured observation. 

On the other hand, you might be more interested in assessing or exploring what 

these phenomena mean – i.e., their qualities – for those in a particular empirical site. 

As John Van Maanen (cited in Alvesson and Deetz 2000:70) suggests, qualitative, 

case study methods like non-structured observation, semi- or unstructured interviews 

and qualitative secondary data encompass “an array of interpretive techniques which 

seek to describe, decode, translate and otherwise come to terms with the meaning, 

not the frequency, of certain more or less naturally occurring phenomena in the social 

world”. So these sorts of methods are better at teasing out qualities as defined above. 

Second, are you inclined towards independence or involvement? Do you think it 

is important to be independent of and detached from your respondents or to be 

involved with them? Do you value objectivity, impersonality and value freedom 

more than subjectivity, interpretation and immersion? If yes, then a quantitative, 

survey method would work best for you. If no, then choose a qualitative, case study 

method. 



Third, do you want to explain or explore? Quantitative, survey methods tend to focus 

on identifying the casual connections or correlations between variables – on trying 

to explain why certain things happen or how they relate to each other. So you might 

want to ask what causes (/produces/generates) motivation, or stress, or productivity in 

a specific organisational environment, or to investigate what lies behind a particular 

market crash. Or perhaps you want to know whether gender relates to earnings in a 

given context, or whether the money supply relates to inflation in another. There is an 

emphasis here on trends and regularities and the discovery of scientific laws about 

human behaviour. 

On the other hand, you might be more interested in exploration. Perhaps you feel 

that looking for causes of or patterns in human behaviour is too mechanistic, 

deterministic, static or reductionist – that it over-simplifies human behaviour and 

ignores our capacity for reflexivity and change, especially as interactions between 

people from different parts of the world are more and more likely as a result of 

globalisation. You would therefore be more interested in exploring a particular sort 

of behaviour as a whole and complex phenomenon instead of trying to ‘boil it down’ 

to its causes or constituent patterns. here there may also be more of an emphasis 

on process and change – on how people come together and influence each other in 

terms of values, beliefs, behaviours etc., and on the idea that, because something 

was once true about human behaviour, this does not mean it will always be. If this 

sounds more like your PMP research then you should choose a qualitative, case study 

method. 

Fourth, do you want to test theories (deductivism) or generate theories 

(inductivism)? Gummesson (2000:63) defines these two approaches as follows: 

“Deductive research starts with existing theories and concepts and 

formulates hypotheses that are subsequently tested; its vantage point 

is received theory. Inductive research starts with real-world data, and 

categories, concepts, patterns, models, and eventually, theories emerge 

from this input.” 

Deductivism therefore involves verifying (or modifying or rejecting) a hypothesis 

about how the world works by testing it against real world data. If you want to do 

deductive research, you should choose a quantitative, survey method. The important 

thing about hypotheses is that they have to be expressed or framed in a way that 

allows them to be tested; so they need to be as precise as possible. Example 4.1 

provides an illustration. 



Example 4.1: Hypothesis testing in deductive research 

Professor X has developed three hypotheses, set out below: 

1. that male undergraduate students of management at the 

University of Leicester watch more television during the average 

week than female undergraduate students of management 

2. that British undergraduate students of management at the 


week than undergraduate students of management from 

elsewhere in the world 

3. that first year undergraduate students of management at the 


week than second and third year undergraduate students of 

management. 

She tests these hypotheses by sending out a self-administered 

questionnaire to all current undergraduate students of management 

about their television viewing habits. Professor X’s data allow her to: 

1. accept hypothesis 1: male undergraduate students of management 

at the University of Leicester DO watch more television during 

the average week than female undergraduate students of 

management, by about 15% 

2. reject hypothesis 2: British undergraduate students of 

management and undergraduate students of management from 

elsewhere in the world at the University of Leicester watch about 

the same amount of television during the average week 

3. modify hypothesis 3: first year undergraduate students of 

management AND second year undergraduate students of 

management at the University of Leicester watch more television 

than third year undergraduate students of management. 

The Oliver et al. (2002) and Martinko and Gardner (1990) papers 

discussed above are also examples of deductive research. 

Inductivism on the other hand gathers the data first and then seeks to understand 

or theorise it afterwards to explain that particular situation. The theory which is 

generated is grounded in these data, and is sometimes claimed to be closer to the 

real world, and more authentic and concrete, as a result. If you want to do inductive 

research, you should choose a qualitative, case study method. 



Fifth, is your research question/s characterised by scope or depth? Research 

characterised by scope is macro research – it looks at the ‘big picture’, and tries to 

identify trends or developments at this level. Econometrics is an example. It uses 

statistical analysis of large samples to identify key factors in the workings of national 

economies and to predict what will happen when these factors change (e.g., the 

interest rate). For macro research, you should choose a quantitative, survey method. 

Research characterised by depth is micro research which focuses more on individuals. 

So a micro economist would be more interested in how actual individuals behave 

in specific economies or economic sectors, and why they behave this way. Micro 

research uses much smaller samples because it aims to know a lot about a small 

number of people (depth) as opposed to knowing a small amount about a lot of 

people (scope/macro). So micro research is more interested in the specific than 

the general (Gummesson 2000:177). Micro researchers sometimes accuse macro 

researchers of producing data and findings which are superficial, overly abstract 

in their statistical correlations and quantifications and “very remote from everyday 

practice” (Alvesson and Deetz 2000:60). If you want to do micro research, you should 

choose a qualitative, case study method. 

But there is still more to choosing an appropriate method/s than this. Let’s move on 

to issues around feasibility and desirability. 

Feasibility and Desirability 

As we have already indicated in our discussion of research questions in Section 3, 

feasibility is very important in research – in other words, is your research actually 

achievable or practical? Do you have sufficient resources in terms of money and 

equipment (e.g., a tape recorder for interviews)? Does the method you have chosen 

reflect your own skills and preferences? For example, if you are very shy, then 

interviewing probably isn’t a great idea. Respondents also react in different ways to 

different methods. The aforementioned task on methods’ strengths and weaknesses 

will introduce you to issues like SAQ fatigue or the intrusive nature of observation 

(which might also affect your ability to gain access to the empirical site). This task 

also establishes that some methods are also more time consuming to administer than 

others – e.g., observation as compared to an SAQ – but then SAQs take a long time to 

design. Further, large samples might not be realistic in terms of how much data they 

generate – do you have time to analyse them all? 

Then there are a whole set of issues around what we might call desirability. here we 

are, firstly, concerned with issues of ethics (e.g., problems associated with covert 

observation – as discussed in the task mentioned above and again in Section 6). 

Second, will the method you have chosen make a contribution (refer back to Sections 

2 and 3)? Finally, will the method you choose allow you to answer your research 

questions? Does it therefore achieve what we might call best fit? In other words, do 

its strengths match your particular research questions? For example, there isn’t much 

point in using structured observation to count the number of incidents involving 

conflict in a particular work environment if what you are interested in is why this 

conflict comes about. Qualitative interviews would be better for this kind of research. 

Equally, if you want to find out about sensitive issues like what causes high levels of 



sickness absence in one particular firm, then face to face interviews might not be the 

ideal way to go. Anonymous SAQs are more likely to encourage respondents to (a) 

take part and (b) be honest! 

Another variation on the methods theme is hybrid methods. We made an implicit 

reference to these above when we were talking about observation potentially being 

both quantitative and qualitative. There we suggested that all research methods can 

be varied to become more or less qualitative or quantitative if circumstances require 

it. An example would be where the ideal choice for a project in terms of best fit is 

semi-structured interviews, but respondents live in a different region or country to 

the researcher and travelling to interview them is not feasible. Thus here an SAQ 

might be used but with a lot of open questions (hence a kind of halfway, qualitative/ 

quantitative hybrid) to gather the qualitative data which the researcher needs. 

Another alternative is to use an Internet chat room or MSN Messenger or Facebook 

to do an interview of sorts in ‘real time’, when again travelling to interact face to face 

with a respondent is not possible. 

This takes us nearly to the end of our discussion of research methods and how to go 

about choosing one. Before we offer our final thoughts, we want to say something 

brief about the use of secondary data, whether qualitative or quantitative, as a 

research method. 

A Note About Using Secondary Data 

As we stated in the introduction, this study book focuses on projects involving the 

collection of primary data. But if you are considering using secondary data, do take 

the following points on board. The list is not exhaustive but it provides you with some 

initial issues to think about. There is also plenty of discussion of the use of secondary 

data in the research methodology texts. See pp. 166–171 of chapter 6 in Blaxter et al. 

(2001) or chapter 21 in Bryman and Bell, for example. 

• Secondary data have been collected by someone else – so it is important 

to be sure that these data will actually enable you to answer your research 

questions, since they were probably collected for different purposes from 

yours. 

• You also need to think about whether the data set you want to use might 

be affected by weaknesses like a non-representative sample (more on 

this in Section 5) or biases like the oft-heard suggestion that governments 

manipulate unemployment figures to paint their administrations in a positive 

light. No data set, primary or secondary, is perfect, but if you use someone 

else’s you do need to be aware of its limitations. In other words, you need to 

find out as much as you can about the methodology and context behind the 

secondary data set you are interested in using, in order that you do not make 

exaggerated claims based on these data in your analysis. 



• how up to date is your secondary data set? Also, is it readily available (e.g., 

already in the public domain)? If it is not, you may have to pay to access it and/ 

or go through a process of gaining permission to use it if it is copyrighted. 

• For secondary qualitative data which document events in the past, are they 

in good condition? Older documents are – obviously – more likely to be 

damaged in some way, and less likely to be archived online. 

• Is the secondary data set raw or has it already been analysed? It is much 

more likely that the latter is the case – in other words, you are not seeing the 

actual answers to the questions set by the original researcher. Instead you get 

their analysis of those raw data. In this instance their interpretation may have 

affected the conclusions in various ways. 

And Finally …. 

We end this section with three important observations about research methods. First, 

no method is more effective across the board than any other. Instead each method 

has its pros and cons. So contrary to popular student belief SAQs are not ‘better’ than 

interviews. Both methods have strengths and weaknesses: the issue is which method 

is most suited to your project. 

Second, flexibility is also paramount. In other words you have to be prepared to 

alter your chosen method (and indeed other aspects of the methodology like sample, 

empirical site etc.) once it’s selected because unforeseen circumstances may make 

it impossible to follow through. For example, you might select semi-structured 

interviews but, when you come to negotiate access with your chosen organisation, 

they will only permit you to send questionnaires to their staff as they are concerned 

that interviews will take up too much of the working day. It’s therefore a good idea 

to have a contingency plan. More about research planning in Section 6. 

Third, remember: Whatever method you choose, you need to be able to justify 

it in your project’s Methodology Chapter! 

Summary 

1. Self-administered questionnaires, structured interviews, observation and 

semi-/unstructured interviews are the methods most commonly used by 

students of management (in our experience at least, which is why we cover 

them here). 

2. All research methods have strengths and weaknesses. 

3. These should be borne in mind when choosing a method, as should issues 

around the kind of research you want to do, feasibility, desirability etc. 



4. You can also use a hybrid method. 

5. You must however be able to defend your choice of method/s in your project. 


� Tasks 









References 




Alvesson, M. and S. Deetz (2000) Doing Critical Management Research London: Sage 

chapter 3 

Baxter, L., C hughes and M. Tight (2001) How To Research 2 nd Edition. Buckingham: 

Open University Press 

Brewis, J. (2000) ‘’When a body meet a body … : experiencing the female body at 

work’ in L. McKie and N. Watson (eds.) Organizing Bodies: Policy, Institutions and 

Work houndmills, Basingstoke: Macmillan pp. 166–184 


University Press chapters 8, 9, 11, 17, 18 and 25 

Cashmore, E. (2002) ‘Behind the window dressing: ethnic minority police perspectives 

on cultural diversity’ Journal of Ethnic and Migration Studies 28(2):327–341 

Collinson, D.L. (1988) ‘“Engineering humour: masculinity, joking and conflict in 

shopfloor relations’ Organization Studies 9(2):181–199 





Gummesson, E. (2000) Qualitative Methods in Management Research 2 nd Edition. 

Thousand Oaks, California: Sage chapters 1 and 3, pp. 172–188 


Learning chapter 9 (NB you may also find useful material in chapters 10, 11 and 12) 

Martinko, M.J. and W. Gardner (1990) ‘Structured observation of managerial work: a 

replication and synthesis’ Journal of Management Studies 27(3): 329-357 

Naudé, P., D. Band, S. Stray and T. Wegner (1997) `An international comparison of 

management’s use of quantitative techniques, and the implications for MBA teaching’ 

Management Learning 28(2):217–233 

Oliver, A., A. Cheyne, J.M. Tomás and S. Cox (2002) ‘The effects of organizational 

and individual factors on occupational accidents’ Journal of Occupational and 

Organizational Psychology 75:473–488 


Practitioner-Researchers 2 nd Edition. Oxford: Blackwell chapters 8, 9 and 11 

Roy, D. (1960) ‘Banana time: job satisfaction and informal interaction’ Human 

Organization 18:158–168 


5 th Edition. harlow: Financial Times Prentice hall chapters 5, 9, 10 and 11 

Silverman, D. (2000) Doing Qualitative Research: A Practical Handbook London: Sage 

chapter 26 

Silverman, D. (2005) Doing Qualitative Research: A Practical Handbook 2 nd Edition. 





section 5 

Sampling, Design and 

Administration




SECTION 5 

Sampling, Design and Administration 



• understand how to gather primary data by sampling 

• understand different types of sampling and the advantages and 

disadvantages of each 

• design your own research schedule according to the method(s) 

chosen to collect your data 

• administer your research schedule effectively. 

Introduction 

Where Section 4 discusses the methods themselves – or at least those most commonly 

used by students like you – and how to select one or more, this section moves to the 

process of primary data collection. Once you have chosen your research method you 

will usually need to develop a sample (i.e., to select your respondents from the whole 

population that you are interested in), to design a research schedule (the list of 

questions you will ask at interview or in an SAQ, or of the issues you want to focus 

on in observation) and to administer it (i.e., to actually gather your data). We discuss 

sampling first. 

Sampling 

Most primary data gathering involves sampling of some kind, because it is usually 

impossible for reasons of time and money to study everyone in the population you 

are interested in. Your population involves everyone that you are interested in (e.g., 

all the employees in a particular organisation; all the distance learning students at a 

particular university; all the nurses at a particular hospital; all the administrative staff 

at a specific city council; all the clients of a local business … and so on). Secondary 

data sets will usually also have been drawn from some kind of sample, as opposed to 

an entire population. There are two basic types of sampling technique; probability 

and non-probability. 


Probability Sampling 


This technique is associated with quantitative, survey research. Probability sampling 

means you can say with confidence that everyone in a given population had an equal 

chance of being included in the sample. Thus the sample is representative of that 

population. In research terms, this also means you are able to generalise from the 

sample to the wider population – i.e., to say that what is happening in the sample 

is also happening in the population from which it is taken. Say for example there 

are 5000 distance learning students at the University of X, and this is the population 

you are interested in. If we take a representative sample of 500 of those students, 

then whatever we find out about these 500 people can also be said to pertain to the 

remaining 4500 people in the population. 

Probability sampling minimises what is known as sampling error. In other words, it 

reduces the likelihood of including non-representative respondents who do not typify 

the characteristics of your population in some way. The highest sampling error which 

is acceptable if you want to claim to have generated a representative sample is 5% 

- i.e., where 5% of your sample do not reflect the population’s characteristics. Put 

another way, this generates a 95% ‘level of certainty’ or confidence level. 

Importantly, probability sampling is only actually possible when you have a complete 

and up-to-date sampling frame. ThIS IS IMPORTANT! A comprehensive sampling 

frame is a full and up to date list of everyone in the population, and Thinking point 

5.1 below asks you to consider this issue in more detail. MOST DIPLOMA STUDENTS 

WILL NOT BE ABLE TO AChIEVE A REPRESENTATIVE SAMPLE BECAUSE ThEY WILL 

NOT hAVE ACCESS TO A COMPLETE SAMPLING FRAME. This does not mean that 

you cannot draw inferences or speculations from your sample, or that it doesn’t 

represent the wider population in which you are interested in any way. But what it 

does mean is that generalisations of the conventional sort are not possible, unless 

your population is relatively small and you have access to a complete sampling frame 

for that population. Also see Section 9 on this issue: students very often make the 

error of suggesting that their samples are representative when they are not, and 

generalising when their data do not in fact permit this kind of analysis, when they 

write their projects up. We also alluded to this kind of error in Section 1. 



Thinking point 5.1: Which of these are complete sampling frames? 

Imagine you were doing research on the following populations, and you 

had access to the associated sampling frames. Which of these sampling 

frames would allow you to develop a representative sample of the 

relevant population? 

1. Population A: all adults living in the city of Leicester 

Sampling frame A: the current electoral register for the city of 

Leicester 

2. Population B: all adults living in the city of Leicester 

Sampling frame B: the current Leicester household telephone 

directory 

3. Population C: all staff at the University of Leicester 

Sampling frame C: the staff list held by the Personnel Department 

at the University 

4. Population D: all staff at the University of Leicester 

Sampling frame D: a list of staff attending the graduation 

ceremonies in January 2010 and July 2010 

5. Population E: all staff at the Leicester branch of the high street 

opticians Specsavers 

Sampling frame E: the payroll list for Leicester Specsavers 

6. Population F: all staff at the Leicester branch of the high street 

opticians Specsavers 

Sampling frame F: a list of staff at the branch compiled by your 

friend who works there part-time 

Four Key Examples of Probability Sampling Techniques 

This is not an exhaustive list! The first and easiest technique is known as simple 

random sampling, which means literally choosing respondents at random from 

your sampling frame until you have the number you need (see below). This can be 

likened to picking people out of a hat; and you could even choose to do it this way. 

Then there is systematic sampling, which involves selecting the first respondent at 

random and then every nth respondent after that, depending on desired sample size. 



So if you had a population of 1000, and you wanted a sample of 100, you would 

choose the first respondent randomly and then every 10th respondent after that. 

Third is stratified random sampling. This is the best way to ensure that your sample 

does in fact reflect your population where you are particularly interested in certain 

demographic characteristics (i.e., the particular strata to which people belong). For 

example, if your population is 40% male and 60% female and you have a specific 

interest in gender as a potentially important variable or factor in your analysis, you 

could divide your sampling frame into two strata of men and women. You would 

then select a random proportionate sample from each stratum 10 to produce an 

overall sample which was also 40% male and 60% female. You can use more than 

one stratum (e.g., gender and occupation), but this makes it complex. The other 

complication with stratified random sampling is that you need a sampling frame 

which gives you more than just a list of names. So, to produce a stratified random 

sample by gender and occupational status, you would need a list that told you who 

was in your population but also their gender and their occupation. 

Our final example is known as cluster sampling. This involves dividing your sampling 

frame into clusters (e.g., by functional area like hRM, Marketing, Finance etc., or by 

organisation) and then selecting the requisite numbers of clusters – the members of 

which will all take part – at random. It is useful for bigger populations. 

In order to be sure that you have a representative sample, you also need to use the 

formula below to identify your sample size. here n a (the end result) is the required 

number of respondents, n is your minimum sample size 11 and re% is expected 

response rate (which will of course be lower for methods like SAQs). 

n a = n x 100 

re% 

(Saunders et al. 2009:221) 

And an illustration, taken from box 7.5 in Saunders et al. (2009:221): 

439 (adjusted minimum sample size for customer survey) x 100 = 

43, 900 

10 The singular form of strata. 

11 See table 7.1 in Saunders et al. (2009:219) for guidance here. As we have 

suggested above, the established confidence level for a representative sample is 95: 

“This means that if your sample was selected 100 times, at least 95 of those samples 

would be certain to represent the characteristics of the population” (Saunders et al. 

2009:218). The table sets out minimum sample sizes for various sizes of population, 

in order to achieve this confidence level. It also maps these against varying margins 

of error, measured in percentages, which again affect the quality of the sample 

drawn. They range from 5% (as suggested, usually the highest margin of error which 

is acceptable in aiming for a representative sample) to 1%. 



Divided by 30 (estimated response rate) = 1463 (actual sample size) 

Non-Probability Sampling 

As you might expect, this type of sampling is non-representative, and therefore does 

not allow for generalisation. It is usually associated with qualitative research where 

researchers tend to be more interested in respondents as individuals as opposed to 

as ‘typical’ members of a wider social group. It might also, as we suggest above, be 

the only option available if you do not have a complete sampling frame. 

Five Key Examples of Non-Probability Sampling Techniques 

Again this is not a comprehensive list, but these are all commonly used approaches. 

The first is quota sampling. This is similar to stratified random sampling in the 

sense that it allows the sample to reflect the population’s characteristics. however 

it does not produce a representative sample because selection of respondents is not 

random, and does not derive from a complete sampling frame. An example would 

be a market survey done in a city centre where a researcher needs 40 men and 60 

women to respond to mirror the city’s gender ratio, but can only approach potential 

respondents who are actually out in the city centre at the time, and passing close 

by to where the researcher is standing. Furthermore, in the end the only people 

who will end up in the actual sample are those who are willing to stop and answer 

questions. Then there is convenience sampling. As its name suggests, this is the 

most straightforward approach where you would select individuals who are easiest 

to access, who are nearest to you and who you know will take part (e.g., colleagues 

or friends or family). 

Third comes purposive sampling which is also known as key informant sampling. 

It involves selecting particular types of respondent because they will be able to 

provide the data that you need to help you to answer your research questions. For 

example, Peters and Waterman (1982) did their research on the characteristics of 

excellent companies by interviewing managers in 43 top performing (i.e., excellent) 

US companies, which were all Fortune 500 listed. They began with a list of 62 of 

McKinsey’s 12 ‘star’ clients and subtracted 19 – including General Electric – on the 

basis of additional performance measures. Those left included hewlett Packard, 

McDonald’s, Procter & Gamble and Disney. This was purposive sampling because 

these managers all worked for excellent companies and so were deemed by Peters 

and Waterman to be qualified to comment on their characteristics. 

Fourth is snowball sampling, which is often used when respondents might be difficult 

to access because you don’t have very much knowledge of the area or where the 

research is controversial or sensitive (refer back to Section 3). It involves identifying 

one or two respondents, hoping they will identify others for you to speak to, who 

will then identify others and so on – so the sample snowballs. And finally there is 

self-selection sampling, where you publicise the research project and ask people to 

12 The management consultancy Peters and Waterman both worked 

for at the time. 



contact you if they are willing to take part. It is often used by those researching for 

television programmes or magazine articles, for example. E-mail distribution lists can 

also be a useful mechanism for producing self-selection samples. 

NB These techniques can also be used in tandem. For example, Brewis (2004) used a 

mixture of convenience, snowball, self-selection and purposive sampling to create her 

focus group sample. This research centred on thirty-something women’s attitudes to 

and experiences of work, personal relationships and urban existence in London. So all 

the women involved had to be around 30 and to have experience of either living or 

working in London (purposive). Brewis approached her existing friendship network to 

generate her sample (convenience and self-selection). Finally, one of her respondents 

was recruited by another woman taking part (snowball). As we have seen, Brewis 

(2000) also used a combination of purposive, convenience and self-selection sampling 

for her research into women’s bodies, as described in Section 4. here she approached 

women (because it was women’s bodies she was interested in; purposive). These 

women were a combination of her existing colleagues and students she already knew 

(convenience), and she sent out a general introductory e-mail to see who was willing 

to participate (self-selection). 

having discussed various ways of selecting the people who take part in your data 

collection and their implications, we now move to designing the actual schedules 

which will structure and inform your interactions with these individuals. 

Designing Research Schedules 

The key issue here is the question “will the evidence and my conclusions stand up 

to the closest scrutiny?” (Raimond, cited in Saunders et al. 2009:156). You need to 

be confident that the claims you make on the basis of your data will stand up to 

examination by your assessors. Asking the right questions and/or focusing on the 

right issues in your data collection will help – and so we neatly segue into the whole 

subject of schedule design. 

To begin with you have a choice: 

1. You can design your own schedule from scratch. 

2. You can use someone else’s, but be careful of copyright issues here – unless 

the schedule is in the public domain you will need to seek formal permission 

to access and use it, and this may mean paying a fee. Plus you will need to 

acknowledge that it is not your original design in your PMP methodology 

chapter, whatever happens. 

3. You can adapt someone else’s, and make the appropriate acknowledgement. 

The choice depends on what best suits your purposes. Our focus here will be on 

designing your own schedule. Clarity of research questions plus subject-specific and 



contextual knowledge are all important in this regard. You need to be sure what it 

is you want to find out about overall (clarity of research questions, as discussed at 

length in Section 3), to be familiar with the relevant academic literature so you can 

explore and assess the relevant concepts, theories, models and findings in your data 

gathering (refer back to Section 2) and to make sure you know as much as possible 

about the empirical context (e.g., a specific organisation) where you are doing the 

data gathering. The last ensures that (a) you look credible to the respondents; (b) the 

questions or issues you are examining make sense to them; and (c) you can verify the 

accuracy of what they tell you if necessary. 

You also need to consider data analysis during the design process. If you don’t think 

about how you will analyse the resultant data while you design your questions/foci, 

how will you be able to make sense of them when collected? Again, more on this in 

Sections 7 and 8. Something else which we would urge you to undertake is a pilot 

test of your schedule once you think it is ready to use. This is especially crucial for 

‘fixed’ designs like SAQs, structured interviews and structured observation where all 

respondents are treated in the same way, but it is arguably important for semi- and 

unstructured interviews and non-structured observation too. A pilot test basically 

means having a trial or practice run with the schedule you have designed to see 

whether it works. Try and use respondents similar to those who will take part in the 

actual data gathering to make the pilot meaningful. You may also want to use any 

data generated from the pilot in your eventual analysis as long as you check this is 

OK with those who participate. 

For methods using direct questions (i.e., SAQs and interviews of whatever sort), your 

pilot test should check the following important issues. When you run the pilot, ask 

your respondents for comments on each: 

• whether they understand the questions and any relevant instructions (more 

of which later) 

• whether the schedule is an appropriate length (overly long SAQs and interviews 

mean respondents will lose patience) 

• whether the questions are arranged in a sensible order, and whether skips/ 

filters work properly (again more on these later) 

• whether any important questions are missing, or if some of the questions you 

have included are irrelevant or redundant 

• whether there are any questions respondents find difficult to answer because 

they are too sensitive or too complex 

and anything else you can think of. 


Once completed, you should also look at the pilot data to see: 


• whether the questions actually generate the data you need to answer your 

research questions (usually referred to as the question of validity) 

• whether some questions seem to result in misinterpretation and the collection 

of irrelevant data (usually referred to as the question of reliability) 

• whether you are going to be able to analyse the data you collect. 

For interviews and observation, a pilot also allows you to practise your interviewing or 

observational skills – which is no bad thing! For structured observation in particular, 

you should pilot in order to ensure that your schedule generates the data you want 

and is straightforward to use. 

Now let’s talk about design itself – i.e., actually putting the schedule together. We 

begin with some general pointers on methods which use direct questions, and then 

move to designing schedules for these specific methods themselves. 

Designing Schedules for Methods Using Direct Questions 

Type of Data 

There are three basic types of data in this regard: 

• factual/attribute/bio-data (information about your respondents’ biographical 

characteristics – e.g., education, age, career history etc.); 

• behavioural data (how they behave in specific situations); 

• attitudinal data (how they feel about specific issues). 

What you ask for depends on what kind of information you need to answer your 

research questions. 

Channel 

You also need to decide on your channel – i.e., the medium you use to actually ask 

the respondents the questions in your schedule. For SAQs you can use snail mail; 

delivery and collection (i.e., you or someone else actually gives the SAQs out by 

hand and then collects them back at a later stage); e-mail; a web-based design; or 

even fax. For interviews the standard choices are face to face or over the phone. 

Sample size and location are both issues here, because the larger and more widely 

spread the sample the more difficult and expensive it can be to reach. So, for a 

large SAQ sample in a number of different regions or countries, say, e-mail or online 



channels are probably more sensible. You also need to ensure that you have the 

contact details you need to operate your chosen channel – postal address, phone 

numbers, e-mail addresses, fax numbers, whatever. Think about resources and 

your skills too. Postal SAQs are costly as we established in the Blackboard task in 

Section 4, but you might not have the equipment or the knowledge to produce 

an online questionnaire. Also consider the respondents’ reaction – e.g., using the 

phone for a structured interview may be too detached and unlikely to encourage 

their commitment or enthusiasm, whereas e-mail or online SAQs could create worries 

about anonymity. Also make life as easy as you can for your respondents, depending 

on who they are, by using the most convenient channel for them. There is no point 

designing an online questionnaire for respondents who do not have regular access 

to the Internet. 

however, there is probably no real substitute for face to face interviewing when you 

are conducting semi- or unstructured interviews as it is essential to build up trust 

and rapport with your respondents. Doing these kinds of interviews over the phone 

would also make the interaction very complex in terms of listening and reacting to 

what the respondent says, as well as recording the interview (if this is an issue). 

Question Types: Closed Versus Open 

Choice of question type is again dependent on the information that you need to 

answer your research questions. The first choice is between closed (where as we have 

seen in Section 5 the respondent’s answer is restricted to a set of given responses) 

and open (where the respondent can answer as they see fit). You can also use a 

mixture of closed and open questions (more on this later). 

There are various types of closed question, some examples of which are as follows: 

• list – here the respondent is asked to choose as many answers as apply to 

them from a list; 

• category – here respondents are asked to choose one answer from a list, 

where categories in the list are mutually exclusive (i.e., do not overlap in any 

way); 

• ranking – here respondents are asked to rank various items in order of 

preference or importance. Ranking questions can be difficult to answer, and 

the number of items should be kept to a minimum. Alternatively, you can ask 

respondents to score each item using a rating or scale question; 

• rating/scale – here respondents are asked to indicate their answer using a 

scale. The classic approach is the 5 point Likert scale (e.g., 1 = strongly 

disagree, 2 = disagree, 3 = neutral/ no opinion, 4 = agree, 5 = strongly agree). 

This kind of question is useful for gathering attitudinal data in particular; 

• quantity – here respondents are asked to indicate the quantity of something 

as it pertains to them; 



• grid – here respondents are asked two (or more questions) at the same time, 

and to fill their answers in on a grid. These can be confusing, and need to be 

clearly worded. 

Example 5.1 provides an illustration of each of these types of closed question. 

Example 5.1: Types of closed question 

List Question 

Please indicate why you have chosen to work for your current employer, 

ticking as many reasons from this list as apply: 

My current employer is the leader in their field 

My current employer pays above the market rate 

My current employer is located near to my home 

My current employer provides flexible working opportunities 

My current employer has a good internal promotion system 

My current employer offers high quality training and development 

Other (please specify) 

Category Question 

how old are you? 

18–25 

26–30 

31–35 

36–40 

41–45 

46–50 

51–60 

61–65 

66 or above 

Did you vote in the 2005 UK General Election? 

Yes 

No 

Can’t remember 



Ranking Question 

Please rank the following from 1 (most important) to 5 (least important) 

in terms of what you look for when purchasing a new cell phone on a 

‘pay as you go’ arrangement 

Cell phone features (e.g., camera, Internet access, e-mail) 

Cell phone design (e.g., flip top, size, colour) 

Price of the cell phone 

Brand (e.g., Sony Ericsson, Samsung, Nokia, Apple) 

Features of the ‘pay as you go’ arrangement (e.g., tariffs, bolt-ons, 

bonuses) 

Rating/Scale Question 

Please indicate how important the following are to you when 

purchasing a new cell phone on a ‘pay as you go’ arrangement. 

Cell phone features (e.g., camera, Internet access, e-mail) 

1 = not at all important, 2 = not very important, 3 = neutral/no 

opinion, 4 = important, 5 = very important 

Cell phone design (e.g., flip top, size, colour) 



Price of the cell phone 



Brand (e.g., Sony Ericsson, Samsung, Nokia, Apple) 



Features of the ‘pay as you go’ arrangement (e.g., tariffs, bolt-ons, 

bonuses) 



Quantity Questions 

how many children do you have? 

how many overseas business trips have you taken in the last twelve 

months? 


Grid Question 

Please complete the following table. 

Names of people who 

live with you 

Their age Their relationship to 

you (e.g., partner, 

daughter, father, 

friend) 


Open questions on the other hand can take the additional form of probes which are 

“worded like open questions but request a particular focus or direction” (Saunders et 

al. 2009:338). Probes can be used as follow up questions to draw a respondent out 

if they have not given sufficient information in response to an initial open question, 

as in Example 5.2. 

Example 5.2: Use of a probe as a follow-up question 

INTERVIEWER: Can you tell me how you felt when your manager 

announced that a new database was being introduced? 

RESPONDENT: Oh, I was worried about it. It took me ages to get used 

to the database we were using and I thought I would have the same 

problem. 

INTERVIEWER (probe): That’s interesting. When you say you took a 

long time to get used to the existing database was this because it was 

complicated to use or was there another reason? 

RESPONDENT: Well database X was very complicated to be honest. I 

didn’t find it very user-friendly and neither did the people I worked with. 

You had to click an awful lot of buttons to get it to co-operate and it 

took a long time to get the answers you needed. But also I just don’t like 

change very much, and I don’t think I’m a very fast learner, especially 

when it comes to IT. I think I’m just too old for it all now [laughs]. 



Probing ensures that you have the amount of detail that you need, but should be 

used carefully (see discussion of ethics in Section 6). 

Question Order and Related Issues 

You should try to make your SAQ or interview schedule interesting and relevant to 

the respondent from the start, so they are motivated to fully take part. Another good 

tip is to warm the respondent up by asking non-controversial questions at the 

beginning before turning to more sensitive issues where required. Similarly, the usual 

recommendation is that you move from straightforward questions to difficult or 

challenging questions which require more thought, again to get the respondent used 

to the data gathering process. Note that attitudinal questions can be the hardest to 

answer. But don’t necessarily leave the more controversial or more difficult questions 

right to the end of the schedule as respondents may be tired by this stage and less 

able or inclined to answer as a result. 

Also on question type, specific questions are much easier to answer than general 

ones – so make it clear to the respondent what it is you are asking about. Plus if you 

are using rating/scale questions, if no ‘neutral/no opinion’ alternative is provided 

in the middle of the scale then respondents who actually are neutral or have no 

opinion will be forced to invent an answer. however, there is also the phenomenon 

of middle drift to consider, where if such an option is provided respondents may 

choose it even if it doesn’t apply to them as they want to avoid seeming too extreme 

in their opinions or behaviours. There are issues of what is called social desirability 

here – i.e., respondents wanting to paint themselves in a good light. Similarly, a 

phenomenon called acquiescence response set apparently means that, if you use 

a closed question with an agree/disagree response set, respondents tend to tick the 

‘agree’ option, regardless of how they actually feel. Again this is because of social 

desirability, as agreeing with a statement can be seen as more socially acceptable 

than disagreeing (depending on the statement of course!). 

Schedule Length 

half an hour is the recommended maximum for phone interviews – although this 

could be seen as rather a lengthy maximum! For face to face interviews length is 

perhaps more negotiable but remember that the longer the interview the more data 

you have to analyse – and the more tired you and your respondent both become. For 

SAQs the shorter the better (as long as you are gathering all the data that you need) 

as again respondents get tired and bored. Saunders et al. stipulate a maximum of 

between 4 and 8 pages for an SAQ, but again 8 pages seems like quite a lot to us. 

You can test this issue out at pilot, as suggested above. 


Translating a Schedule 


You may have to translate your schedule (and indeed the resultant data) from the 

language spoken by the respondents into English for the project itself 13 . So you need 

to think about what might be ‘lost in translation’. Saunders et al. (2009:383–384) 

offer some helpful tips in this regard. One of these is the importance of considering 

what they call experiential meaning. The term ‘dual career household’ (i.e., where 

both parents work outside of the home), say, may not be understood if respondents 

have no experience of such a phenomenon. So this would not be an appropriate 

term to use in such a context. Another example is the difference between idiomatic 

meanings and literal or lexical meanings – so ‘grapevine’ in English means the 

same, idiomatically or colloquially (i.e., in English slang), as ‘téléphone arabe’ in 

French. That is to say, both terms refer to a gossip network. Neither, however, imply 

their literal meanings (something which grapes grow on, for the English version, or 

an Arab telephone in the French!). 

What to Avoid 

Lengthy or imprecise questions confuse respondents and lead to poor quality data 

as the respondent will not be sure what is being asked. Multiple questions (i.e., 

asking two or three questions in one) also tend to mean that only the last one will be 

answered. For example, avoid asking questions like ‘When were you last appraised 

and to what extent do you feel you benefited from it?’ Leading questions imply the 

answer in the question – e.g., ‘When did you last socialise with your colleagues?’ 

assumes that the respondent does socialise with colleagues. This means that the 

respondent will probably answer in a specific way. Again questions like these are to 

be avoided. 

Also try not to use technical language and/or abbreviations that respondents may 

not be familiar with. Insensitive terminology is likewise to be avoided. For example, 

don’t refer to a ‘coloured person’: this is politically incorrect. Use ‘person of colour’ 

instead. Further, don’t use negative questions. Asking something like ‘I have never 

experienced sexual harassment in the workplace: agree/disagree’ may produce double 

negative answers like ‘I disagree that I have never experienced sexual harassment in 

the workplace’. Such questions are confusing for the respondent and for you. Finally, 

don’t ask questions that look as if you are examining respondents on their knowledge 

of specific issues as this can make them feel uncomfortable and resentful. 

having established some of the generic issues relating to the design of direct 

questions, we now discuss what you need to bear in mind when designing methodspecific 

schedules. First we look at SAQs and structured interviews. 

13 Obviously this also applies if you want preliminary comments on your schedule 

from a Blackboard project tutor (see Section 9). 



Designing Self-Administered Questionnaires and 

Structured Interview Schedules 

Obviously these two methods are similar in the sense that they involve a predetermined 

list of questions, but they differ in that SAQs are self-administered whereas the 

interviewer asks the questions and records the answers in a structured interview. 

Closed or Open Questions? 

Ordinarily, as we have established in Section 4, you would tend to keep open 

questions to a minimum in SAQs and structured interviews, first because they are 

more difficult for respondents to answer in this sort of format, and second because 

researchers using these methods usually aim for a standardised approach where every 

respondent is confined to the same response sets. But you should always consider 

adding an ‘other: please specify’ option to a list of fixed responses in case the 

respondent’s own situation isn’t covered, and/or providing space for them to expand 

on a response to a closed question if they wish in an SAQ. Also open questions can 

be used in these sorts of schedules, especially if you are using a hybrid type method 

as also discussed in Section 4. 

Start with your Research Question/s 

You should start the design process with your overall research question and 

subquestions and subdivide these into themes. For example, if the research 

question(/s) has to do with how respondents feel about the recently introduced 

legislation banning smoking in enclosed UK workplaces, then the data needed might 

include their own smoking behaviour plus their attitudes about passive smoking 

(inhaling other people’s tobacco smoke) and whether smokers should be provided 

with a separate smoking room inside an organisation instead of being compelled to 

smoke outside. You might also want to know whether such behaviours and attitudes 

seem to relate to respondents’ age or gender. Then repeat this step to check that your 

proposed themes are exhaustive enough to gather all the data you need. 

Think About Data Type and Measurement Issues 

having done this, identify the type of data needed under each theme. For example, 

the data described above for the research on the workplace smoking ban in the 

UK fall into all three categories – attitudinal (feelings about passive smoking and 

smoking rooms), behavioural (does the respondent smoke?) and factual/attribute/ 

bio-data (age and gender). So think about the best kinds of questions for the types 

of data needed. Then you need to consider appropriate measurements. Level of 

detail is the issue here. For the example above, you would need to know how many 

cigarettes respondents smoke a day as well as simply asking whether they smoke, as 

this is likely to have some relationship to their attitudes to the ban itself. 


What Should the Schedule itself Contain? How Should it be Laid Out? 


Obviously for SAQs the schedule will be sent to the respondent in writing, whether 

this is distribution and collection, snail mail, e-mail, online or by fax. For structured 

interviews the researcher will read material out to the respondent from the schedule. 

For SAQs in particular then, but arguably structured interviews too, choose a clear 

title which grabs respondents’ interest. Then, in the introduction to either type of 

schedule (which is crucial), you need to tell the respondents the following: 

• why they have been selected for the research (e.g., because they have a 

specific sort of occupation) and how they were selected (e.g., randomly from 

a list of a particular population) 

• who you are and what the purpose of your research is (i.e., your Professional 

Management Project, as part of your Professional Diploma in Management 

programme at the University of Leicester School of Management) 

• what the research is about (make this fairly brief) 

• that confidentiality and anonymity 14 will be maintained, if appropriate – 

and this is also the point at which you would ask interview respondents if you 

can record the interview, if this is what you want to do 

• roughly how long the SAQ or interview will take to complete. 

At this stage you might want to ask interview respondents if they have any questions, 

or provide your contact details for SAQ respondents for the same reason. For SAQs 

you also need to provide instructions for answering the questions (e.g., answer as 

many as apply to you, tick the relevant responses, etc.). 

Much of the above could be done in the form of a ‘warning letter’ sent to respondents 

a week or so before the interview is done or the SAQ is sent. Some researchers do 

both – i.e., repeat the warning letter on the SAQ or as an introduction to the interview 

itself. Note that instructions need to be clear for all question types and respondents 

may need reminders as the schedule progresses. Interestingly, for closed question 

14 Confidentiality as we define it means that only you see or hear the raw data 

– i.e., listen to the tapes of the interviews or read the notes you have taken, or 

have access to completed questionnaires, data collected using structured observation 

schedules or non-structured observational field notes. It can also be interpreted to 

mean that you will not share the information generated by the data gathering (i.e., 

your analysis) with anyone other than your assessors, so no one else has access to the 

final project. Remember that having someone else transcribe (more of which later) 

interview tapes/recordings for you immediately breaches both kinds of confidentiality, 

so don’t promise this if you are going to have your data transcribed in this way. 

Anonymity means that you disguise the identities of the organisations and individuals 

concerned. Indeed with SAQs in particular you may never know who the respondents 

are. We will come back to these issues in Section 6. 



response sets, research suggests that people prefer ticking boxes to circling answers, 

so bear this in mind. 

You also need to think about the positioning of bio-data questions. As we have seen, 

these are factual questions about who the respondent is – like their gender, age, 

income etc. They can act as a warm-up as they are easy to answer. On the other 

hand, questions about age and income might be off-putting at the start as some 

respondents could see these as quite sensitive. We suggest in any case that you 

ask for data about income or age using bands (eg £25, 000–£29, 999; £30, 000– 

£34, 999; £35, 000–£39, 999 etc. OR 18–25, 26–30, 31–35 etc.) rather than directly. 

Also please be sure that you need these data – don’t ask for them just because 

everyone else seems to! If you don’t have any particular use for information about 

respondents’ age, gender, income etc. then don’t include these questions. 

You should think about dividing your schedule into sub-sections, because this can 

make it seem shorter and more manageable to respondents. Under each section the 

questions should then be the same for everyone in the sample. Obviously questions 

should be read out verbatim for structured interviews. You may also want to use 

flash cards with longer response sets printed on them for specific questions. In 

this way you don’t have to read out lists of alternatives which the respondent then 

has to remember. Instead you can simply pass them the flash card after posing the 

question. And for both SAQS and structured interviews, specified prompts can help 

a respondent to understand open questions – e.g., you might ask ‘how would you 

describe your manager’s style?’ then the prompts would be ‘democratic, authoritarian, 

has an open door-policy, hands-off, hands-on etc.’. 

Most if not all SAQ/structured interview schedules also contain what are known 

as filters or skips. These are used when questions may not apply to every single 

respondent. For example, you might ask whether a respondent has children. If they 

do they should then answer the questions that follow on childcare. If they don’t they 

can be directed to move straight to a later section in an SAQ or the interviewer would 

skip to these later questions in an interview format. Filters/skips need to be carefully 

designed so that they work properly! This again is something you should check at the 

pilot stage. 

Finally, in terms of your closing comments, say thank you, ask respondents 

whether there is anything else they would like to add and for SAQs include return 

instructions (which you may also want to add to the introduction, and repeat at 

the end). You might offer to send them a copy (or executive summary) of the data 

analysis if appropriate. This might be simply a matter of ethics or courtesy, or it could 

be because you would like their comments on a draft version of the analysis. This 

approach is called respondent validation or member check and can help to ensure 

that your analysis actually reflects what the respondents have told you. More on the 

ethics of respondent validation in Section 6. 


Designing Semi-Structured and 

Unstructured Interview Schedules 


Again we take these two methods together as they are so closely related. There are 

many similarities here to the process already outlined in the section above, but we 

highlight differences where they exist. 

Closed or Open Questions? 

The majority of questions in these sorts of schedules will probably be open with a 

view to gathering rich, qualitative data. however, closed questions can be used; for 

example, for gathering bio-/factual data and warming the respondent up. Probes are 

also particularly useful in this sort of interviewing. 

Follow the Advice Given Already 

As suggested, the process of designing schedules for these sorts of interviews is 

similar to that for SAQ/structured interview design. But remember that most of the 

questions will be open and that the actual questions you ask and the order in which 

you ask them will vary depending on the respondent – in other words, your schedule 

is indicative only. It will give you a list of themes that you want to cover and some 

suggestions for relevant questions under each. 

What Should the Schedule itself Contain? How Should it be Laid Out? 

Begin with an introductory spiel about who you are, why you are doing the research, 

why you want to interview the respondent, why and how they were selected and 

what the research is for, as well as its main focus. Offer reassurances of confidentiality 

and anonymity (if appropriate), and ask whether you can record the interview (if 

appropriate). You should also ask whether the respondent has any questions for you. 

Again all of this could also be done via a ‘warning letter’ as above. The schedule 

should then continue into a list of headings with indicative questions under each. In 

other words, the headings cover the broad issues you want to explore plus suggested 

wordings for the questions which come under each heading. Also think about probes 

– what kinds of questions you might use to get respondents to say more about 

specific subjects. Your closing comments should again take the form of a thank you, 

plus asking the respondent whether they have anything else to add and offering 

them a copy of your data analysis if appropriate (as above). Example 5.3 contains an 

excerpt from a focus group schedule to illustrate what these kinds of schedules might 

look like. 



Example 5.3: Excerpt from a focus group schedule 

What follows is an excerpt from the schedule that Brewis (2004) used 

to gather data from a series of focus groups. As we saw earlier in this 

section, her research questions focused on thirty-something women’s 

attitudes to and experiences of work, personal relationships and urban 

existence in London. here are some of her indicative questions. 

• What would be your perfect job? has this ideal changed? If you 

don’t have your perfect job, why not? Does such a thing exist? 

• Do you consider yourself to be good at your job? Why? What 

does it mean to be successful at work? 

• What does work mean to you? Is it important for women to 

work? Would you ever consider relying on a partner for financial 

support? have any of these opinions changed over the course of 

time? 

• Would you say you have made sacrifices/put other things on hold 

for your job? If so, what and why? If not, why not? 

• What comes first, work or personal life? has this always been 

true? Is work/life balance important to you? Why/not? Do you 

think you have achieved it? 

• What does it mean to be in a relationship? Is it possible to have 

the perfect relationship/to find one’s other half? What about 

monogamy? 

• Why do you live/work in London or the environs? Does it fulfil your 

expectations? Would you move somewhere else if you could? If 

so, where and why don’t you? If not, why not? 

• What would you change about living here? What are the key 

urban problems? how do they affect you? 



Just before we move to look at designing observation schedules, please also bear 

these wise words in mind: 

“The researcher should be warned against assuming that a ‘nondirective’ 

interview, where the interviewee talks freely without 

interruption or intervention, is the way to achieve a clear picture of 

the interviewee’s perspective. This is far from true. It is more likely 

to produce no clear picture in the mind of the interviewee of what 

questions or issues the interviewer is interested in, and in the mind 

of the interviewer, of what questions the interviewee is answering.” 

(Easterby-Smith et al. 2008:142) 

In other words, unstructured, non-directive interviews can be confusing and unsettling 

for the respondent; a bit like a therapeutic interview where the therapist is relatively 

silent and expects the interviewee to do all the talking. They won’t necessarily allow 

you to actually get a sense of respondents’ thoughts and opinions – so make sure, if 

you want to do this kind of interview, that you are clear on your purpose and areas 

of interest. It’s certainly not a method to use when you don’t really know what you 

want to look at! 

We now leave behind schedules which use direct questions and move to briefly 

consider issues around the design of observation schedules, where you are watching 

what your respondents do as opposed to asking them questions about their behaviour. 

Designing Observation Schedules 

Data to be gathered here might relate to: 

• the layout of the space you are observing 

• the actors within it and what they do 

• the objects in the space (e.g., desks, chairs, PCs), and how the actors ‘interact’ 

with these objects 

• specific occasions like meetings 

• time issues (e.g., how long it takes to do something, how frequently things 

are done or the sequence in which they are done) 

• goals (what do the actors’ goals appear to be?); and feelings (likewise). 

Remember also that behaviour can be verbal (what is said), non-verbal (body 

movements, gestures, expressions etc.), spatial (how people move around an area), 

extra-linguistic (e.g., tone, volume, pace of speech) etc. Obviously you need to 

think about collecting the kind of data which will allow you to answer your research 



questions. So even with non-structured observation there will be issues you are more 

– or less – interested in. 

For structured observation, however, you need to consider the following issues when 

designing your schedule: 

• Are the categories on your schedule clearly defined so that it will be easy to 

use in practice? 

• You need to ensure that your categories don’t overlap, and that they cover 

all the issues you are interested in – you may also have to include a ‘residual’ 

category for behaviours that aren’t easy to categorise. 

• Is the schedule easy to use, like a simple tick box or tally mark scheme? 

Now we have covered sampling and design, let’s consider what you should be aware 

of when you are actually ‘in the field’; i.e., collecting your primary data/administering 

your research schedule. 

Administering Research Schedules 

Administering a Questionnaire 

First, make sure your schedule is presented in either 10, 11 or 12 point text so that it 

is easy for respondents to read. Twelve point should always be the default setting, as 

it is so standard. Also try and strike a balance between presenting the questionnaire 

on too many pages so that it looks lengthy and off-putting, and squashing it all in so 

it becomes difficult to read. Third, you might want to consider using coloured paper 

if you are printing the questionnaire out so that it grabs attention – e.g., yellow 

and pink are both said to be warm and appealing colours. Or use colours and/or 

different fonts and typefaces for instructions or subtitles to make them stand out. 

however, don’t overdo it: this will make the SAQ look untidy and unprofessional. And 

remember that capitals and italics can also be hard to read. Further, we suggest you 

stick to commonly used fonts like Arial or Times New Roman. 

If using snail mail or distribution and collection as your channel, make sure the SAQ 

is printed on good quality paper, and use good quality envelopes. This is a matter of 

professionalism and making the right first impression. Also if using snail mail, firstclass 

post or the equivalent gets the SAQ there faster and enhances first impressions: 

the same is true of using correct names and addresses on the envelopes. Also include 

a stamped or franked self-addressed envelope if possible to encourage response – 

although, as with first class post, this does increase your costs. In terms of when you 

should send out/distribute an SAQ, if they are going to a workplace, the usual advice 

is to time them to arrive at the start of the working week. If you are sending SAQs to 

private homes (unlikely in your project research, and may be viewed as an invasion of 



privacy), time them to arrive at the start of the weekend. Also think about the time of 

year – avoid major vacations (e.g., Christmas in the UK) as far as you can. 

You can also use follow-ups to increase response rate. This means sending a reminder 

and/or a duplicate questionnaire to respondents after the deadline has passed, to 

encourage those who haven’t responded to do so. This would of course need to 

go to everyone if it is an anonymous SAQ as you won’t know who has responded 

already. If this is the case, remember to thank those who have responded in your 

introduction/covering letter. The usual advice is to send the first follow-up a week 

after the deadline has passed. Up to three follow-ups have been shown to improve 

response rates, but there are time and resource implications here, and in any case this 

finding might not apply to all samples. Remember also that research using SAQs is 

usually survey type research, which implies treating all respondents identically in your 

administration to achieve standardisation. 

Administering Interviews – General Points 

We start with some generic observations about conducting interviews before offering 

some more specific pointers about different types of interview. First then, in terms of 

arranging when the interviews will take place, try and accommodate your respondents 

rather than them having to accommodate you when setting interviews up. This is 

just a matter of common courtesy. And be honest about how long the interview is 

likely to take so respondents can schedule accordingly: you should know this from 

your pilot test. Plus be sensitive to logistics when arranging interviews. Make sure 

you have enough time to travel from one to another, and leave yourself space to 

have breaks and so on. And think about the setting for the interview. For example, 

interviewing people at work about work may cause them to be reticent for fear of 

being overheard – so might another setting be more appropriate? On the other hand, 

asking them to come on to ‘your territory’ might also unsettle respondents. Consider 

the best location for your data gathering carefully. For example, when Brewis (2004) 

ran the aforementioned focus groups, she held them in her own home: because she 

already knew the majority of the women taking part, this was a familiar, comfortable 

and welcoming setting for the data gathering. 

Before you start the interview proper, use some icebreakers – i.e., innocuous questions 

about their day so far – to help to warm respondents up and to establish rapport. 

You can also collect bio-data early on as already suggested as a warm-up, as well as 

these data being useful information in and of themselves (bearing in mind though 

that some bio-data may be regarded as sensitive). Throughout the interview try as 

far as you can not to give away your own perceptions and biases too much through 

your non-verbal cues (body language, expressions, tone of voice, emphasis etc.), 

and try not to interrupt when respondents are speaking. Show that you are actively 

listening – maintain an attentive posture and an interested tone of voice as well as 

a natural, conversational level of eye contact. Moreover, going too slowly will make 

the interview boring for both of you, whereas going too fast means the respondent 

doesn’t have time to consider or expand on their answers. 



Recording the data themselves is also an issue. Audio recording of whatever kind 

allows you to focus on asking questions and listening to respondents. It also gives 

you a full record of the data including ‘ums’, ‘ahs’, pauses, laughter, tone etc., and 

it means you can use direct quotations from the data in your analysis. On the other 

hand, these benefits may be less of an issue in structured interviews, and using a 

digital or tape recorder may also mean that the respondent focuses on the machine 

and feels inhibited. In addition, you may experience technical difficulties (e.g. tapes 

snapping, batteries running out and technical malfunctions). Always take written 

notes anyway, or make a written record straight after the interview, and don’t forget 

to ask for the respondent’s permission to record before the interview starts. You 

could even give the respondent control over the recorder – i.e., allow them to turn 

it off when they don’t want to be recorded talking about a specific issue. We will 

discuss whether or not you need a transcript of your data (i.e., a full and complete 

typed record of everything that was said) in Section 8. And on a related note, what 

of off the record or hand on the door data? Sometimes respondents will tell you 

things after the recorder has been turned off at the end of the session or you’ve 

stopped taking notes … what do you do with these data? It’s up to you, but you 

must never use them if respondents specifically request that you do not. 

Administering a Structured Interview 

As well as all the pointers above, there are some specific issues concerning the 

administration of this type of interview. First, print the schedule on one side of the 

paper only so you don’t have to flip pages back and forward when asking questions. 

Make sure too that the instructions you need to follow are clearly indicated so you can 

follow them easily – e.g., when to skip a question and so on. As for your demeanour, 

in general a pleasant but professional approach is probably called for here. A fairly 

businesslike outfit is also recommended as the general idea is usually to treat all 

respondents the same and to maintain some kind of objectivity or distance from 

them. 

Don’t try and learn your schedule off by heart – this creates too much pressure on 

you – but try and be fairly familiar with it so you can maintain a natural level of 

eye contact with the respondent, and not have to keep looking down. And keep to 

the schedule: after all, you chose this method for its structure! Make sure that you 

record all answers accurately, perhaps with a tape/digital recorder, or by making the 

majority of your questions closed (as already discussed) so you can just tick boxes. 

And remember the issue of identical treatment, as discussed above for SAQs. This is 

harder to maintain for structured interviews as they take place for the most part face 

to face, and variations are therefore more difficult to avoid. 

Administering Semi- and Unstructured Interviews 

In this case it can sometimes be an idea to send respondents a list of themes or 

aims of the interview in advance (with your ‘warning letter’ maybe?) so they can 

give these some thought before the actual session takes place. Appearance-wise, we 

recommend that you tailor your appearance to them. If they are likely to be smart, you 

should dress smartly and vice versa. Remember that the usual idea here to set them 



at their ease and maximise rapport as a result. During the interview itself you need 

to listen carefully to what respondents are saying and how they are saying it so you 

can decide on the next question quickly as well as to prevent respondents becoming 

upset by a line of questioning (more on which in Section 6). Make sure you cover all 

the relevant areas before closing an interview – check your schedule! If you are not 

sure what a respondent means, ask them to clarify using a probe. Summarising at 

certain junctures what you have picked up from what a respondent has told you also 

ensures you have understood what they are saying. Finally, again being relatively 

familiar with your schedule is a good idea for reasons discussed above. 

Administering Observation 

Pointers here include maintaining your chosen type of observation (e.g., structured 

and non-participant). Presumably you have decided on a particular approach for 

good reasons so stick to it unless there are overwhelming reasons to change. For 

non-structured observation, take notes on the spot as far as possible. Use written 

notes or a laptop or palmtop, so you are recording things whilst they are happening 

and not relying on your memory, unless you have agreed to video what is going on 

as well. And don’t start another session until you’ve written up your data from the 

previous observation session – otherwise sessions can become blurred in your mind. 

You may also need to let respondents get used to you being there for a while – so 

that you become ‘part of the furniture’ – before actually starting to record the data 

proper. This can reduce the observer effect, where your presence makes people 

behave in unnatural ways. 

Summary 

1. Sampling techniques determine your ability to generalise; but it is highly 

unlikely that you will be able to achieve a representative sample in your PMP. 

2. A good schedule design means your findings will stand up to the ‘closest 

scrutiny’. 

3. Attention to a range of issues is paramount here – e.g., pilot test, data type 

and detail, closed versus open questions, choice of channel, length of schedule 

etc. 

4. Administering a schedule also requires consideration of several factors; e.g., 

whether to use follow-ups for SAQs, logistics for setting up interviews, only 

printing the schedule on one side of the paper for structured interviews etc. 




� Tasks 









References 




Alvesson, M. and S. Deetz (2000) Doing Critical Management Research London: Sage 

chapter 3 

Brewis, J. (2004) ‘Sex and not the city? The aspirations of the thirty-something 

working woman’ Urban Studies 41(9):1821–1838 

Brewis, J. (2000) ‘When a body meet a body ... : experiencing the female body at 

work’ in L. McKie and N. Watson (eds.) Organising Bodies: Policy, Institutions and 

Work houndmills, Basingstoke: Macmillan pp. 166–184 


University Press chapters 7, 8, 9, 10, 11, 17 and 18 



Gummesson, E. (2000) Qualitative Methods in Management Research 2 nd Edition. 

Thousand Oaks, California: Sage chapters 1 and 3, pp. 172–188 


Learning chapters 4 and 8 

Peters, T.J. and R.h. Waterman Jr. (1982) In Search of Excellence: Lessons From 

America’s Best-Run Companies New York: harper and Row 




Practitioner-Researchers 2 nd Edition. Oxford: Blackwell chapters 8, 9 and 11 

Saunders, M., P. Lewis and A. Thornhill (2009) Research Methods for Business 

Students. 5 th Edition. harlow: Financial Times Prentice hall chapters 7, 9, 10 and 11 

Silverman, D. (2000) Doing Qualitative Research: A Practical Handbook London: Sage 

chapter 26 

Silverman, D. (2005) Doing Qualitative Research: A Practical Handbook, 2 nd Edition. 





section 6 

Planning, Access and Ethics




SECTION 6 

Planning, Access and Ethics 



• plan your research within a flexible time structure 

• consider how to gain access to data ‘in the field’ 

• understand the difference between physical and cognitive access 

• understand the importance and value of maintaining an ethical 

approach 

• understand how to design your project so it conforms to ethical 

standards. 

Introduction 

In this section we cover issues around planning and timetabling your research, getting 

access to empirical sites to collect primary data and the broad area of research 

ethics. Please note that research ethics in particular is an area that we would urge 

you to take very seriously indeed: it is impossible to cover all of the relevant themes 

in a book like this, so you are encouraged to read more widely around this subject 

especially. 

Research Planning 

Make a Plan But Don’t Expect it to Work Out this Way! 

“Trouble awaits those unwary souls who believe that research flows 

smoothly and naturally from questions to answers via a well-organized 

data collection system” (hodgson and Rollnick, cited in Robson 2002:84) 

Planning your research – i.e., creating a step by step timetable for everything you 

need to do up to and including submission of your project – allows you to structure 

your time effectively and to cover all the necessary bases. But bear hodgson and 

Rollnick’s warning in mind, and make your plan flexible. New opportunities can 



come up unexpectedly (like the publication of a new piece of research which suggests 

a slightly different direction to the one you originally selected for your project). The 

same is true of unforeseen circumstances (like not achieving access to your chosen 

organisation/s or a change in your personal situation such as a new job). Build in as 

much slack time – i.e., time to deal with such opportunities and circumstances – as 

you can. We also recommend that you have contingency plans as already suggested 

in Section 4 – e.g., a backup list of other organisations which you can contact for 

access if your first choice(/s) doesn’t work out. 

Another thing to remember is that research almost always takes longer than you 

think it will. hodgson and Rollnick provide a series of half-humorous and half-serious 

pointers in this regard – e.g., that getting started on the project itself will take twice 

as long as the data collection and that a research project will change at least twice 

in the middle of the process. Plans should also work concurrently as far as possible. 

In other words, you can be working on several steps at once, such as drafting a 

literature review at the same time as negotiating research access, or working on your 

data analysis at the same time as drafting your methodology chapter. 

The Desirable Versus the Possible 

“In the conflict between the desirable and the possible [in research], 

the possible always wins.” 

(Buchanan et al., cited in Saunders et al. 2009:71) 

Our second epigram in this section is one of the most sensible things ever said about 

research! What you actually achieve (the possible) is almost always different from 

what you set out to achieve (the desirable). This is not problematic in terms of the 

quality of the finished project as long as (a) you have slack time and contingency 

plans in place when the desired approach turns out not to be feasible; and (b) you 

acknowledge any limitations of your research due to the difference between the 

desirable and the possible in the appropriate places in the project. Refer back to 

Section 1 for more discussion of the latter point. 

What Should a Research Plan Contain? 

As distance learning students you will all be completing your projects over different 

periods of time depending on your individual circumstances. You should therefore 

start the planning process by setting a date when you would like to finally submit. 

Don’t make this the same date as the date when you must submit the project in order 

to gain your degree (i.e., two years after your initial registration). Leave yourself some 

time in between – again to allow for any opportunities or problems to be dealt with. 

Also remember, as set out in the Diploma handbook, that all your coursework, as well 

as your project, must be completed by 1st September if you wish to be considered 

at the November/December Examination Board and to have the opportunity 

to graduate the following January. Likewise,1st March is the deadline for those 

who wish to be considered at the summer Examination Board, so as to have the 

opportunity to graduate in July that year. BUT DO NOT TRY TO RUSh ThROUGh YOUR 

PROJECT! MANY STUDENTS IN ThE PAST hAVE TRIED TO SQUEEZE ThEIR RESEARCh 



INTO AN INAPPROPRIATE TIMEFRAME, JUST TO SUBMIT EIThER BY 1ST SEPTEMBER 

OR 1ST MARCh. ThEY hAVE FAILED ThEIR PROJECT MODULE AS A RESULT. DON’T 

LET ThIS hAPPEN TO YOU. IT IS MUCh MORE SENSIBLE TO TAKE ThE TIME ThAT YOU 

NEED TO DO ThE BEST JOB YOU CAN! 

Once you have decided your submission date, you can then work backwards from 

this date and allocate weeks or months when each step of the plan will be undertaken. 

What follows are the elements of a plan for empirical projects. For theoretical or 

library projects, the steps relating to methodological design, gaining access and data 

gathering and analysis should be replaced by additional plans for accessing, reading, 

evaluating and synthesising subject-specific literature/s. 

• Give yourself time to generate your topic in the first instance, and to 

identify your research question/s. As we have suggested in Sections 2 and 

3, developing research questions should involve a scan of the subject-specific 

literature. This material – on organisational structure, organisational culture, 

relationship marketing, charismatic leadership, stress, productivity, employee 

motivation, the supply chain, auditing, whatever – will form the basis for 

refining your research ideas. 

• Ensure that your timetable makes room for ongoing discussions with your 

supervisor via one of the Discussion Boards on Blackboard (see Section 9). 

This should allow for discussion before the tutor approves your proposal (see 

below) and afterwards while the research proper is taking place. 

• Decide when you will produce your draft literature review, remembering 

that you also need to make time for keeping up with any developments or 

‘new finds’ in the subject-specific literature throughout the project process. 

The literature review has many functions, as noted in Section 2, but one 

additional benefit of putting together a draft is that it gets you over the 

first hurdle of starting to write the actual project. Also consider where you 

are likely to get your literature from – your distance learning centre?; local 

university libraries?; through the full text databases on the University of 

Leicester library website?; elsewhere on the Internet?; all of the above? etc. 

There are of also course time implications relating to each of these options. If 

you want to use material from other university libraries you need to check out 

what your access and borrowing privileges will be – if any. Students located 

in the UK might also want to consider using the British Library in London (see 

http://www.bl.uk/ for details). Refer back to your Foundations of Management 

module book for tips on where you can find the relevant material (Section 4: 

Learning Resources). 

• Also give yourself time to (re)read the methodology literature in order 

to finalise your methodology (how you will go about gathering/accessing 

your empirical data, whether they are primary or secondary). If you are doing 

primary empirical work then deciding on your methodology means identifying 



an appropriate method but also choosing your sample and how you will go 

about selecting it. You will also need to design your schedule, if appropriate, 

and pilot it. 

• You need to write your one page research proposal. See the Proposal 

Proforma – the structure of which must be followed for all proposal 

submissions – in the same place as the Project Guidelines on Blackboard. 

The Guidelines also offer additional support on how to compile the proposal. 

The proposal should be submitted via the Discussion Board you are using, 

once both you and the tutor there are satisfied that you have identified 

suitable research questions, suitable subject-specific literature/s and a suitable 

methodology where relevant. We also strongly recommend that you continue 

with your literature review and/or your methodological reading whilst you 

are waiting for formal tutor approval as opposed to embarking on your 

data collection, in case the tutor recommends substantial changes to your 

research ideas. PLEASE ALSO NOTE ThAT, ALThOUGh ThE PROPOSAL DOES 

NOT FORM PART OF ThE MARK EVENTUALLY AWARDED FOR ThE PROJECT, 

YOU MUST hAVE SUBMITTED A PROPOSAL, AND hAVE hAD IT APPROVED ON 

BLACKBOARD. OThERWISE WE WILL NOT BE ABLE TO MARK YOUR PROJECT 

WhEN IT IS SUBMITTED. YOU MUST ALSO DOWNLOAD AND INCLUDE ThE 

APPROVAL MESSAGE FROM ThE TUTOR AS PART OF YOUR FINAL PROJECT 

SUBMISSION. PROJECTS SUBMITTED WIThOUT ThIS APPROVAL MESSAGE 

WILL BE RETURNED WIThOUT hAVING BEEN ASSESSED. AGAIN, SEE ThE 

PROJECT GUIDELINES FOR MORE INFORMATION. 

• Do you need to gain research access? If so, when will you do this? Also build 

in time to alter your methodology or to secure access elsewhere if necessary 

following these negotiations. More on access below. 

• When will you actually collect/access your primary and/or secondary data? 

When will you produce your draft methodology chapter? When will you 

analyse your data and draft the relevant chapter? All of these stages need 

to be included in your plan. 

• Also make time for producing your introduction, conclusion, 

recommendations, reflections, bibliography, appendices and redrafts of 

chapters already written. 

• Finally, what about formatting, proofreading etc.? Again see the Project 

Guidelines), and make time for all of this in your plan. 

• Which leads us to your final submission date. 

We also recommend that, for empirical projects, you keep a diary of how your data 

gathering progresses from start to finish (from choice of method through to data 

analysis). This will help as an aide mémoire when you are writing up your methodology 

and data analysis chapters. Something else we recommend when devising your plan 

is to check it out with one of the project support forum tutors on Blackboard at an 



early stage, to ensure it is both feasible and comprehensive. Also see Thinking point 

6.1 for an illustration of how not to do it! 

Thinking point 6.1: A very ambitious research plan 

Giulia is a Professional Diploma in Management student. She also 

manages a local clothes shop, and often works six days a week, as well 

as having a partner, three children and an active social life. Giulia always 

feels as if she is short of time, and tends to try and do her Diploma work 

as quickly as she can because she has so many other things demanding 

her attention. Giulia has decided that for her project she wants to find 

out about the extent to which the independent shops in her local area 

have training and development plans in place, how staff feel about the 

presence or absence of these plans and whether this has a negative 

or positive impact on their behaviour at work. This is as a result of her 

own experiences of managing the clothes shop, and also because of 

the reading she has done which suggests that small and medium sized 

businesses – and retail businesses in particular – tend to neglect formal 

hRM activities. The literature argues that this can have a detrimental 

effect on staff’s motivation, commitment and performance. 

Giulia wants to interview four shop managers and four members of 

staff (one per shop). But she has left her project very late and now has 

only six weeks until her registration on the programme expires, on 1st 

September 2010. She has finished reading the Professional Management 

Project module book and develops the following research plan to try 

and cram all the necessary work in: 

Generate topic/research questions: as a result of her reading for her 

Managing People and Organizations elective module, Giulia has already 

decided on her topic and refined her research questions. 

Draft literature review: Giulia sets aside one week for reading and 

drafting her literature review. She knows that she can’t just resubmit 

her MPO assignment as the literature review. However she thinks she 

can return to the notes she took for this assignment, and do a very fast 

search on the University databases to update these notes, then turn it 

all into the finished draft in seven days. 15.7.10–21.7.10 

Read methodology literature and finalise methodology: Giulia allows a 

week for this, including designing and piloting her interview schedules 

(one for managers, one for staff). She is going to use her deputy manager 

and one of her staff team to pilot the schedules. 22.7.10–28.7.10 



Writing and submitting one page project proposal: Giulia decides she 

doesn’t have time to do this, so leaves it out of the plan altogether. 

Gaining access: Giulia already has good contacts with her fellow local 

shop managers so she reckons she can make a few quick phone calls 

one evening and set all of the interviews up that way. 29.7.10 

Collect data: Giulia sets aside two weeks for this. She knows that retail 

staff often work part time so aren’t always available on any given day. 

30.7.10–13.8.10 

Draft methodology: Another week is set aside here. Giulia thinks she’s 

done the necessary reading already so putting the draft together 

shouldn’t take long. 14.8.10–21.8.10 

Data analysis and draft data analysis chapter: Giulia figures she will 

have all the material she needs by this point – both subject-specific 

literature and data – so she sets aside another week for this process. 

22.8.10–28.8.10 

Introduction, conclusion, recommendations, reflections, bibliography, 

appendices, redraft: Giulia decides she can do all of this in two days. 

29.8.10–30.8.10 

Formatting and proofreading: Giulia sets aside one day for this. 31.8.10 

Final submission date: 1.9.2010 

Giulia has made several fundamental errors in putting her research plan 

together. Based on our recommendations for compiling a research plan, 

what do you think they are? What can you do to avoid them in your 

own plan? 

And so to actually getting in to organisations – or wherever – to collect your primary 

data. In other words, we now consider the vexed question of achieving access. 

Research Access 

How Much Access Do You Need? What About Cognitive Access? 

We know from a Section 4 task that some methods are more intrusive than others – 

e.g., observation (more intrusive, especially in its participant form) versus SAQs (less 

intrusive, especially where you never meet the respondents). There are also different 

types of access. Physical access means getting agreement to actually gather your 



data from the gatekeeper 15 of the organisation/s concerned. Cognitive access 

means being able to access your respondents’ real opinions – getting them to trust 

you and open up to you. So while you may achieve physical access – say the hR 

Director gives you permission to interview a sample of the employees at Organisation 

X – you also need to consider how you will go about achieving cognitive access so 

that your respondents will actually provide you with meaningful data. Techniques 

like the ‘warning letter’, good schedule design and piloting, effective administration 

(all covered in Section 5) and ensuring that you pay careful attention to the ethics of 

research (covered later in this section) help in this latter respect. 

Leave Enough Time 

Negotiating access can be time consuming especially if you are cold-calling (i.e., 

you have no prior contact with or knowledge of the organisation). Then you need 

to make the initial contact, and possibly follow it up if you don’t hear anything for a 

while. You may also not select the right person to approach in the first instance and 

have to start again with someone else. Further, you may need to physically meet with 

someone at the organisation/s to agree the terms of your access. These difficulties 

and obstacles are all good reasons why access negotiation should not be ignored 

in your research plan, and why you need to set aside a sensible amount of time to 

complete the negotiations. 

Warm Contacts Versus Cold-Calling 

The usual advice is to draw on any track record you have with the organisation/s 

where you want to collect or access data – e.g., a partner, relative or friend who 

works there or has worked there in the past. These kinds of warm contacts allow 

those inside the organisation to vouch for you, tell you whom to speak to and so on 

– they generally smooth the path of access negotiations. Indeed perhaps you work at 

the organisation now or have worked there previously (i.e., you are an insider). This 

is beneficial as you will have some knowledge of the organisation’s history and its 

politics, as well as who to approach for permission. Plus it should mean that you have 

considerable credibility with the people concerned. On the downside, perhaps there 

are or were status differences between you and your respondents – could this create 

problems? how objective can you be about such a familiar setting, if this is important 

to you? If you are still working there, how easy will it be to keep data confidential if 

necessary? Plus if you make mistakes in gathering your data, will you then have to live 

with them at work? And so on … 

If cold-calling, whom should you contact? We recommend that you start with the 

most senior member of the human Resources team or its equivalent – the hR manager 

or director – or alternatively the head of the Public Relations department if this seems 

more appropriate. You could even try the CEO if it is a small organisation. These 

senior members of the organisation would usually act as gatekeepers, brokers or 

patrons for your research. In other words, they may be able to give broad agreement 

15 This is the person who has authority to grant you permission to do your 

research at that specific site. 



for access, but might then pass you on to someone else to negotiate the details. If 

you don’t know who the relevant individual is for your organisation/s, you could 

contact their telephone switchboard or their Internet site for details. Alternatively, 

you could try finding out from the local Chamber of Commerce or its equivalent, 

a professional association (e.g., the UK’s Chartered Institute of Personnel and 

Development), a trade union etc. Above all, make sure you get the person’s title, 

name and postal or e-mail address right! 

If cold-calling, write in the first instance – possibly by e-mail, but a hard copy letter 

is more formal and professional. however, e-mail may speed things up especially if 

the organisation/s concerned are in another region or country to you. The letter itself 

should: 

1. Tell the recipient about your research topic and why you are interested in their 

organisation. 

2. Establish your credibility (say that you are a Professional Diploma in 

Management distance learning student at the University of Leicester). 

3. Indicate the kind of access you would like, including method, sample size, the 

time and any resources (e.g., an interview room) you will need and when you 

would ideally like to do the data collection/access. 

4. State anything you can offer in return because “the more the company 

gives, in time or money, the more it expects in exchange” (Easterby-Smith et 

al. 2008:30). You could for example offer some kind of report on your data 

to managers and/respondents. Maybe your research focuses on a topic of 

particular significance to the organisation which otherwise they would not 

have the resources or time to explore. But do make sure whatever you offer 

is actually going to happen! 

5. What can you guarantee? Perhaps that you will make minimal use of their 

time and resources? Perhaps that you can offer anonymity and confidentiality? 

6. Use formal language and the business letter format. 

7. Write clearly and keep it as short as possible – the person you are writing to 

is likely to be very busy. 

8. Make replying to your initial request easy – include a stamped or franked selfaddressed 

envelope/e-mail address/fax number/(cell/mobile?) phone number. 

You could even use a pro forma where the individual can simply tick ‘yes’ or 

‘no’ to your request, and if it is a ‘yes’ indicate times and dates to discuss this 

further if required, or give details of another person to speak to if appropriate. 



9. If you are sending a hard copy letter, word process or type it, address it 

personally, sign it by hand and print the envelope if possible. Follow the letter 

up with a phone call if you don’t hear anything but wait for at least 2–3 

weeks to do this! 

Negotiating the Contract and What Happens Afterwards 

Drawing up a written contract regarding your research access might be a good idea, 

especially in a cold-calling situation, as then both sides will be sure about what has 

been agreed. But whatever happens the final details of your research access need to 

be spelled out carefully and clearly. Also be careful not to give away too much when 

negotiating access. For example, if the organisation demands proofreading rights 

this could create a lot of difficulties for you at a later date if they require changes that 

detract from the substance of your argument. Alternatively, the organisation may ask 

you to sign a confidentiality agreement. We strongly recommend that you discuss 

this with a tutor on the appropriate Blackboard support forum before signing any 

such document, to ensure that it will not create any problems in terms of assessment. 

Thereafter, you may want to consider keeping a contact report – i.e., a diary noting 

when agreed actions were completed – and sending regular updates to the relevant 

organisational personnel so everyone is informed about your progress. 

As suggested above, once physical access has been agreed, you may want to contact 

respondents separately via a ‘warning letter’ to enhance your chances of achieving 

cognitive access. Finally, honesty is imperative throughout regarding what you would 

like in terms of access and what you can do in return. 

And finally in this section, one of the trickiest issues of all … 

Research Ethics: The Basics 

At the risk of sounding repetitive, this subject is extremely important. ULSM tutors 

will take ethical considerations into account when assessing a project proposal and 

of course the eventual submission itself, as we feel that this is a vital aspect of any 

sound piece of research. We therefore strongly recommend that you spend some 

time reading in more detail about research ethics as what we cover here is only the 

basics 16 . 

You should also note that, as part of your one page research proposal, you will need 

to answer specific questions about the ethics of your proposed research. Please see 

the Project Guidelines on Blackboard for more about the ethics approval process. 

16 Of course this is true of all the other issues covered in this book, even when 

taken in conjunction with the companion textbook, the Project Guidelines and the 

other resources available on Blackboard. But we particularly stress the need to read 

more widely regarding research ethics because it is a central part of a good project 

and yet tends in our experience to be neglected by students. 


however, for the time being you need to remember two key issues: 


1. No field research with live human beings can be undertaken by any 

student or member of staff at the University of Leicester without receiving 

prior ethical approval. In other words, if you are doing any kind of primary 

empirical research for your project, then the Blackboard tutor will take this 

into account when assessing your proposal. Approval of the proposal signifies 

ethical approval in all cases. Where ethical approval is not given you will be 

required to submit a revised proposal. 

2. Please also be aware that, if you answer the questions about ethics on 

your proposal incorrectly, then you will not obtain ethical approval for your 

research. Again this will require resubmission of the proposal. 

But what do we actually mean by research ethics? Ethics in academic research can 

be roughly defined as “the appropriateness of your behaviour in relation to the 

rights of those who become the subject of your [research] work, or are affected 

by it” (Saunders et al. 2009:183–184). Remember that certain people may be 

affected by your research even if they don’t participate in it in any way – e.g., if your 

recommendations are implemented by the organisation/s concerned. As a general 

rule of thumb, researchers should 

“safeguard the proper interests of those involved in or affected by their 

work … ensure that the physical, social and psychological well-being of 

research participants is not adversely affected by the research … [and] 

attempt to anticipate, and to guard against, consequences for research 

participants that can be predicted to be harmful.” (British Sociological 

Association 2002: paragraphs 6, 13 and 26) 

You might find it useful to consult the BSA’S Statement of Ethical Practice yourself 

for more information on ethics. Similar sorts of statements can be found in other 

scholarly associations’ codes of practice – e.g., the British Psychological Society, 

the Academy of Management, the American Psychological Association, the American 

Sociological Association etc. You need to consider research ethics throughout any 

empirical project – when considering your topic and questions, choosing your 

method, selecting your sample, designing your schedule, gaining access, collecting 

data, analysing them and writing up the project itself. 

It is also worth bearing in mind Fine et al.’s (2003:178) observation that “ethnography 

depends upon human relationships, engagement and attachment, with the research 

process potentially placing research subjects at grave risk of manipulation and 

betrayal”. Although Fine et al. refer specifically to ethnography here (as discussed in 

Section 4), we suggest the use of qualitative methods in general probably exacerbates 

the need to be alert to research ethics. This is because qualitative methods usually 

allow us to get closer to our respondents, to find out more about them – and so 

they become more identifiable in the final write-up and more vulnerable as a result. 

This does not however mean that those conducting quantitative research can ignore 

ethics – they are very important in this regard as well! 



What follows is a discussion of what we would regard as the basic ways in which 

you can try to ensure that your research is as ethical as possible. Research ethics as a 

subject area goes well beyond the issues discussed here, but this material should give 

you an idea of what we regard as the minimum. 

Securing Informed Consent 

Informed consent means that those who take part in the research do so of their 

own free will, without any coercion or deception by the researcher, and have enough 

information to be able to make that decision sensibly. It requires that you spell out 

the aims and nature of the research, who is doing it and what the results will be 

used for to potential respondents – and also that you allow them to choose whether 

they do in fact participate. Informed consent may even involve generating a written 

contract which both you and each respondent should sign. But, whether written 

down or not, the following are all important aspects of securing informed consent: 

• respondents should know who you are and how you can be contacted; 

• respondents should understand what the research topic is and why the 

research is being carried out. They should also understand what is being 

asked of them and how the resulting data will be used; 

• respondents should be assured of confidentiality and anonymity if appropriate; 

• respondents should be told that it is their choice as to whether they take part; 

• respondents should be told that they are free to withdraw from the research 

process at any time, and that they don’t have to answer all your questions; 

• respondents should be asked if they will consent to the reproduction of any 

direct quotations from what they have told you; 

• if the informed consent contract is written down, then it should be signed 

and dated by both you and the respondent to confirm their agreement and 

they should be given a copy. 

Ensuring informed consent can take place in conjunction with the ‘warning letter’ 

discussed in Section 5 or during the introduction to the interview/SAQ/observation 

session itself. 

Unfortunately many researchers have not followed this basic ethical rule in the past. 

Indeed in 1985, Adair et al. (cited in Robson 2002:69) suggested that “upwards of 

81% of studies published in the top social psychological journals use deception in their 

procedures”. Examples include Roy’s participant observation in a factory (see Section 

4), where his co-workers did not know he was an academic. The defence for this kind 

of approach is usually that it minimises what we have referred to elsewhere as the 

observer effect, so that socially acceptable answers or behaviours from informants 



are not an issue. however, it is still highly unethical as it does not give participants 

any choice about whether they actually take part in the research. 

A different example is Milgram’s (1974) infamous experiments on the connection 

between authority and conformity. These, very briefly, consisted of pairs of respondents 

who were in separate rooms. One of each pair (the ‘teacher’) asked the other (the 

‘learner’) questions. Every time a question was answered incorrectly, the teacher had 

to inflict an electric shock on the learner. These shocks ranged up to 240 volts, which 

was designated as a ‘dangerous’ level. The teachers were told by Milgram that the 

experiment was in fact about the relationship between pain and learning – i.e., does 

inflicting electric shocks help someone to learn information more effectively? But in 

reality Milgram wanted to see whether, if one was told to obey an order (authority), 

would the order be followed even if it was unpleasant (conformity)? So the actual 

subjects of the experiment were not the learners, but the teachers, although the 

latter were completely unaware of this fact. 

The authority–conformity relationship was assessed via one of Milgram’s team circling 

the experimental site to tell any teacher who began to show disquiet at having to 

inflict increasing levels of shocks that ‘The experiment requires you to continue’. 

Things were made worse by learners audibly screaming in pain, calling for mercy 

or, even more dramatically, falling silent altogether. The experiment actually showed 

that the majority of the teachers did inflict the maximum voltage, despite a group of 

psychiatric experts being asked beforehand what would happen and predicting that 

only a tiny percentage would in fact go this far. 

You will be relieved to know that in fact no electric shocks were involved. These were 

faked, as were the reactions of the learners. Nonetheless, the teachers in this case had 

not given their informed consent as they were unaware of the true purpose of the 

research; and so the research can be considered very problematic in this regard. We 

should also consider covert note-taking and recording in this context. Furthermore, 

remember that the standard line in methodological discussions in the contemporary 

social sciences is that covert research of any kind is not acceptable. 

Also problematic in relation to informed consent is the tactic of coercion. Robson 

talks, for example, about prisoners being offered early release or extra food for 

taking part in trials of potentially dangerous drugs. Coercion though can relate to 

offering any kind of incentive to potential respondents to participate. Does ‘bribing’ 

respondents in this way mean that they have in fact given their full consent? This is 

debatable. 

On a different note, one of the authors of this book conducted interviews in a university 

and a financial services organisation for her PhD research. When she approached 

staff at the university to ask them to take part, she told them she had already gained 

permission from the head of their department to do so. Although this was true, and 

she passed on this information so that the potential respondents would know she 

had official permission to do her research, it seemed some felt as a result that they 

had to take part. In other words, certain people interpreted the head of department’s 



permission as an implicit order for them to agree to be interviewed. This could also 

be regarded as an instance of (unintended) coercion. 

Not Placing Respondents Under Any Physical or Psychological Duress 

The Milgram experiment has also been widely criticised for subjecting its participants 

to duress – in other words, making them undergo a potentially harmful or stressful 

experience. Similarly, Zimbardo et al.’s (1973) research created a mock prison 

environment, using paid volunteers, to explore how people respond to the roles they 

are given. Do they fulfil the role as it is ‘officially’ stated or do they negotiate and 

interpret it themselves? In other words, do our roles shape us or do we shape our 

roles? The young male participants were designated as either prisoners or guards, 

and given the appropriate clothes and props. The experiment was supposed to 

last for a fortnight but was abandoned after six days. This was because the guards 

were often cruel and tried to break the prisoners’ spirits. The prisoners, for their 

part, became servile, felt dehumanised and depressed, had fits of crying and rage, 

and exhibited disorganised thinking. Some had nervous breakdowns. Nearly all the 

prisoners begged to be released early and were happy to forfeit any payment. In 

short, it seems from this experiment that it is roles that shape us, not the other 

way round. But the ethical question here is whether it was appropriate to test this 

hypothesis using these means. Furthermore, could Zimbardo and colleagues not have 

predicted the unpleasant psychological outcomes which resulted for the men who 

participated? 17 

Respondent Validation/Member Check 

We have referred to this idea before in Section 5. We can understand the practice 

of asking respondents to comment on draft data analyses as allowing you and them 

to move towards a common understanding of the research site and thus enhancing 

the credibility or validity of the findings. But doing a ‘member check’ also has ethical 

ramifications because it offers respondents the courtesy of being able to approve the 

ways in which they are represented in the eventual research text. 

17 It is worth noting here that Milgram’s discussions with eminent psychiatrists 

about the apparently very small likelihood that any teacher would go to the maximum 

voltage is something he raises as a defence of his approach. Likewise, Zimbardo and 

colleagues had screened all applicants for the prison experiment so as to select only 

those who appeared intelligent, stable, well adjusted and generally normal – thus 

(they thought) minimising the likelihood of any psychological repercussions for those 

who took part. 


Maintaining Confidentiality and Anonymity 

As discussed in Section 5 and above. 

Reporting Genuine Data 


Transparency throughout the research process is essential. You should seek to ensure 

that you are able to provide evidence of your research process and data collection, 

as advised in Section 1. It is good practice to keep and store diaries, correspondence 

with anyone at your empirical site/s, completed questionnaires, video/audio tapes 

of all interviews, photographs etc. ULSM markers have in the past asked students 

to provide evidence of their data collection following submission of their projects, 

where doubts have arisen as to the accuracy of their claims about their data or indeed 

whether their data exist at all. Your data can then be destroyed once you have passed 

your project and achieved your Diploma. 

Again there are many examples of researchers having falsified data in the past. Broad 

and Wade (1983) describe several instances of this kind. One famous case is British 

psychologist Sir Cyril Burt’s statistical studies of the intelligence of identical twins. 

he claimed to have found that their IQs were identical even when they had been 

adopted by different families at birth and thus brought up apart. In other words, 

the key assertions here are that intelligence is (a) genetically inherited and (b) fixed 

throughout life. This research was taken so seriously that, after World War II, Burt 

became involved in establishing a new two-tier system of secondary education in the 

UK based on his findings. Children took an examination called the 11 plus and were 

then placed into grammar schools (for the more academically gifted) or secondary 

moderns (for those deemed to be more suited to a vocational education). But in 

the 1970s, after his death, Burt was publicly accused of having fabricated his data. 

For example, in three separate studies of different numbers of identical twins, he 

reported the same statistical correlation of IQ scores to the third decimal point. This is 

an incredible, not to say highly implausible, finding. Similar flaws exist in his research 

reports as far back as 1909. It has also been suggested that his two field investigators/ 

co-authors, ‘Margaret howard’ and ‘J. Conway’ never actually existed. 

Tensions and Ambiguities in Research Ethics 

What follows is a discussion of some of the ways in which it can be difficult or 

challenging to put even this ‘bare minimum’ of ethical considerations into practice. 

This is not intended to deter you from taking ethics very seriously indeed. Instead it 

should alert you to some of the ways in which you might breach ethical rules or face 

ethical challenges during the research process, even if you are trying to do the ‘right 

thing’. 



Informed Consent: Observation and the ‘Public Interest Defence’ 

The first question here is whether informed consent is something that researchers 

should focus on only at the beginning of the data collection process, which then 

satisfies this particular ethical requirement. This is especially relevant to observation. 

We have talked in Section 5 about allowing respondents to get used to your presence 

as an observer in their daily habitat and for you to become ‘part of the furniture’. But 

do we also have the responsibility to continually remind people of whom we are and 

why we are in their organisation? Otherwise they may forget that they are talking 

to or behaving in front of a researcher and so let their guard down unwittingly. 

The British Sociological Association (2002:paragraph 25) actually say that informed 

consent may need to be an ongoing practice in situations of this kind. 

The second question is the so-called ‘public interest defence’. Milgram, for example, 

has justified his research protocol on the basis that he was examining a key social 

issue, based on the ‘I was only obeying orders’ explanation from the guards of the 

Nazi SS after their atrocities were revealed in the wake of World War II. Milgram 

suggests that his approach was intended to assess whether ‘normal’ people in ‘normal’ 

situations actually obeyed even unpleasant or unacceptable orders in the same way, 

and so his results were therefore of significant public interest. A related example is a 

recent television documentary, aired in the UK, where a journalist went undercover 

with the extreme right wing British National Party and filmed their activities and 

opinions without those involved being aware he was a reporter. The documentary 

itself was extremely shocking, and could again be defended on the basis that the 

‘real’ BNP attitudes and behaviours would not have come to light in any other way. 

But the British Sociological Association again suggest that any covert research needs 

a very robust and substantial defence of this kind, and even then is not immune from 

ethical criticisms. 

What Does It Actually Mean To Not Place Respondents Under Duress? 

We have already discussed some fairly glaring examples of placing respondents under 

duress in the shape of the Milgram and Zimbardo et al. experiments. But subjecting 

respondents to some form of duress doesn’t have to be as dramatic as these examples. 

It can be as simple as requiring them to answer all the questions in an interview, 

or probing for clarification or more details beyond their comfort zone. There are 

issues here of the initial choice of research topic/question/s as well as how sensitive 

we are to the ways in which respondents react to our questions. Reactions aren’t just 

verbal either. Facial expressions, gestures, body posture, nervous laughter etc. might 

all signal that a respondent is having difficulty discussing a specific issue. 

For example, Ramazanoğlu and holland (1994) cite Kelly discussing her research into 

adult female survivors of sexual abuse, and how she had to bring some interviews 

to a premature close because the respondents became so upset. This might seem 

like a pretty obvious example because the subject matter here is extremely personal 

and intimate. however, Ramazanoğlu and holland, while suggesting that their own 

research into young women’s perceptions of sexual risk and AIDS also falls into 

this category, go on to argue that they were faced with an additional challenge 



during this project. Some of their respondents struggled with vocabulary during the 

interviews because the ‘scientific’ terms for certain body parts – like ‘penis’ or ‘vagina’ 

– which the interviewers were using weren’t ones which these young women used as 

a matter of course. however, the respondents also worried that their own preferred 

terminology for these body parts was too crude and embarrassing to use in the 

context of an interview. Reassuring the young women in this respect required what 

Silverman and Perakyla (cited in Ramazanoğlu and holland 1994:139) call “elegant 

interactional work” by the interviewers. 

The Realities of Respondent Validation/Member Check 

As holliday, following Skeggs, notes, “discursive frameworks make interpretation 

from a critical perspective impossible for some respondents. The … practice of having 

respondents confirm one’s interpretation is by no means a guarantee of ‘truth’” 

(2000:518). If we consider the power relations at work here and the status that 

academic researchers still enjoy in many societies, it becomes clear that respondents 

might indeed find it difficult to challenge a representation that an academic has 

made of them, even if they object to it. In other words, maybe they feel that the 

academic is the expert and knows them better than they know themselves. This may 

not be as much of an issue for you as Diploma students but it could arise if you are 

doing research in your own organisation with more junior members of staff. 

When And How Should We Maintain Anonymity? 

hearn (2004) documents his experiences of discrimination as a non-Finn applying 

for a professorial post at a Finnish university. he refers to how his academic standing 

and qualifications were misrepresented. hearn also says that being able to speak 

Finnish had been explicitly signalled in the advertisement and further particulars for 

the post as irrelevant, but then reappeared as a criterion for his eventual rejection. 

The process lasted from August 1998 when the advertisement appeared to April 

2001 when another candidate was confirmed as having been successful. Despite the 

fact that hearn’s version of events is highly critical of those involved, no attempt is 

made to anonymise this series of events. All individuals and institutions are named, 

and chunks of correspondence between hearn and other players are reproduced 

verbatim. 

hearn notes that that the Finnish system of academic recruitment is totally open in 

the sense that a list of all applicants for a job is made public, as are referees’ reports 

as well as documents on how the faculty reached their eventual decision. he says he 

did consider disguising identities – including his own – but decided this would be 

“disingenuous” given the public nature of many of the relevant documents and the 

fact that the case had been widely reported in the Finnish media (hearn 2004:41). 

hearn’s paper therefore raises two key ethical questions. First, do we owe any duty 

of anonymity to respondents whose behaviour we see as problematic? Second, in 

qualitative research especially, we need to bear in mind that full anonymity often 

requires more than just changing the names of people, organisations and places. 

We may also need to attend to job titles, ages, genders, ethnicities, educational 

qualifications, country of location, even language used in order to make sure that 



nothing and no one is identifiable from the text. But is it actually possible to fully 

anonymise qualitative data in this way without sacrificing their complexity, nuances, 

specificity to their locale, how they report respondents’ words and constructions – 

and thus their impact? After all, the richness and detail of qualitative data are the 

main reasons why some researchers choose this approach in the first place. 

Also 

“Suppose that, [in the course of your data gathering] in an office, 

school or hospital setting, you observe serious and persistent bullying by 

someone in a position of power; or that people are being put at physical 

or other risk by someone’s dereliction of duty.” (Robson 2002:71) 

What do you do in this instance if you have also promised anonymity to the bully – how 

can you report their behaviour? You should in the first instance post a query to the 

relevant Blackboard project support forum for advice from a ULSM tutor (obviously 

without giving anything away about who or which organisation is involved). They 

may recommend that you take it up formally with the organisation regardless of the 

anonymity guarantee, or withdraw from involvement with the people concerned. 

Even more difficult is the instance where something like this occurs in your own 

organisation. 

In all, these questions do not have easy answers – but we raise them because you 

need to be aware of them from the start of your project research. 

Summary 

1. Research plans covering the whole project process are useful but should 

contain as much slack as possible, as well as contingency plans. 

2. Gaining access requires leaving enough time, talking to the right people, 

making viable promises, devising a clear contract and being honest. 

3. Also remember that being an insider has benefits and drawbacks in this 

respect, and that there is a difference between physical and cognitive access. 

4. There are ‘bare minimum’ ethical standards in research which need to be 

considered throughout … 

5. ... but these involve tensions and ambiguities in themselves: for example, 

should you breach someone’s anonymity if you hear about or see them doing 

something you consider unacceptable? 



� Tasks 










References 




British Sociological Association (2002) Statement of Ethical Practice Available online 

at: http://www.britsoc.co.uk/equality/Statement+Ethical+Practice.htm (accessed 5 th 

November 2009) 

Broad, W. and N. Wade (1983) Betrayers of the Truth New York: Simon and Schuster 


London: Sage pp. 129–138 

Fine, M., L. Weis, S. Weseen and L. Wong (2003) ‘For whom? Qualitative research, 

representations and social responsibilities’ in N.K. Denzin and Y.S. Lincoln (eds.) The 

Landscape of Qualitative Research 2 nd Edition. Thousand Oaks, California: Sage pp. 

167–207 

hearn, J. (2004) ‘Personal resistance through persistence to organizational resistance 

through distance’ in R. Thomas, A. Mills and J.h. Mills (eds.) Identity Politics at Work: 

Resisting Gender, Gendering Resistance Abingdon, Oxfordshire: Routledge pp. 40–63 

holliday, R. (2000) ‘We’ve been framed: visualising methodology’ Sociological Review 

48(4):503–521 

Milgram, S. (1974) Obedience to Authority London: Tavistock Publications (also see 

Milgram, S. (1994) ‘Conformity and independence’ in h. Clark, J. Chandler and J. Barry 

(eds.) Organization and Identities: Text and Readings in Organizational Behaviour 

London: Chapman and hall, pp. 132–144) 



Ramazanoğlu, C. and J. holland (1994) ‘Coming to conclusions: power and 

interpretation in researching young women’s sexuality’, in M. Maynard and J. Purvis 

(eds.) Researching Women’s Lives from a Feminist Perspective London: Taylor and 

Francis pp. 125–148 


Practitioner-Researchers 2 nd Edition. Oxford: Blackwell pp. 65–76, pp. 376–384 



Roy, D.F. (1960) ‘Banana time: job satisfaction and informal interaction’ Human 

Organization 18:156–168 

Zimbardo, P.G., C. haney, W.C. Banks and D. Jaffe (1973) ‘The mind is a formidable 

jailor: “a Pirandellian prison’’’ New York Times Magazine 8 th April pp. 38–60 






section 7 

Analysing Quantitative Data




SECTION 7 

Analysing Quantitative Data 



• outline different types of statistics and different types of 

quantitative data 

• discuss different ways in which quantitative data can be measured 

• identify various graphical and numerical representations of 

descriptive statistics 

• appreciate the role of inference in statistical analysis. 

Introduction 

This section is concerned with techniques for analysing quantitative data. It approaches 

this broad subject area from the perspective of those of you who need to understand 

how other people’s data have been manipulated to produce their research findings, 

as well as those who want to analyse their own quantitative data. If you have studied 

Quantitative Analysis for Management as an elective module, some of this material 

will already be familiar to you. We cover types of statistics and types of data, discuss 

scales of measurement for quantitative data of different types, explore the difference 

between graphical and numerical representations of descriptive data and discuss the 

two approaches to making predictions based on our observations using inferential 

statistics. 

The Terminology of Statistics 

We all use statistics on a regular basis, usually without knowing it, including those 

of us who insist that we are not numerically minded. Statistics help us to make sense 

of our experiences. We might observe, for example, that our commute to work in the 

morning takes us between 60 minutes and 85 minutes depending upon the traffic 

and the time we leave home. In discussions with colleagues we may come to the 

conclusion, based on our shared observations, that congestion in our local area is 

increasing, and it is therefore taking us longer to get to work. We might ponder how 



long it will take us next year to travel from home to work. We might even decide that 

we want to find a new job closer to home as a result of all these considerations. 

In statistical terms, here we are generalising from what we have observed. We can 

also distinguish between two types of statistics that are apparent in this example. 

In observing the time it takes us to commute to work we are providing descriptive 

statistics’ i.e., a description of our observation. Within the example we are also 

inferring (from the observations we have made) an observation that we cannot 

actually make because we don’t have direct experience of it, namely that congestion 

in the whole area is increasing. We may wish to refer to this as inferential statistics, 

i.e. making estimates or predictions based on our observations. 

Descriptive Data 

Descriptive data can be complex when there is a large group of elements which we 

want to understand. We might therefore want to summarise the descriptive data into 

tables, graphs or simply numbers. As we have seen in Section 5, we may be interested 

in such a large group of elements (individuals, companies, voters, households, 

products, customers and so on) that time, cost and other considerations mean we 

only actually gather data about a small portion of the whole group. The whole group, 

as you will remember, is the population, and the smaller group, a subset of the 

population, is our sample. 

To provide an example, imagine that our earlier observations about congestion 

were viewed seriously enough by local government to require a broader study. The 

population might be defined as individuals travelling within the area during peak 

times. We would not have sufficient resources to study each of these individuals and 

so we would seek to produce a sample of the population. If this was representative, 

we would seek to infer the behaviour of the population as a whole from this sample. 

As we established in Section 5, researchers who work with quantitative data usually 

do want to establish representative, probability samples. 

Quantitative data is the term used to identify the facts and figures that are collected, 

analysed and summarised in any statistical investigation. The group of data collected 

to address a particular situation is often referred to as the data set for the study, 

although this term is used for qualitative data as well. Within a quantitative data 

set are a range of elements that comprise the subject of the study. For example, the 

data set may contain a list of the students registered on the Professional Diploma 

in Management programme. For each student the data set may contain important 

information such as age, gender, work experience and so on. These characteristics 

make up the different variables to be investigated. Data is obtained by collecting 

measurements for the specific variable in question. The measurement for the variable 

collected from a particular element (here each student) is known as the observation. 

In our example, the details of Paula Jones, an imaginary Diploma student who is 32, 

would be the observation of the variable of age. 



Both qualitative and quantitative data can be classified as being either cross-sectional 

or time series 18 . Cross-sectional data is collected at, or pertains to, a particular point 

in time; for example, the profitability of a number of companies in the 2009–2010 

financial year, or the expenditure of a group of professionals on personal computing 

equipment in a given month or year. In contrast, time series data refers to information 

collected over, or pertaining to, a period of time. If we stay with the examples just 

mentioned, we could examine the profitability of a number of companies every 

year during a five-year period, or we could examine the expenditure of a group of 

professionals on personal computing equipment each quarter over a period of two 

years. 

Scales of Measurement 

The approach adopted in the statistical analysis of data on a particular variable will 

depend upon the scale of the measurement used for that variable. There are four 

possible scales of measurement; nominal, ordinal, interval and ratio. The type of scale 

used will determine how the data can be manipulated. A variable is being measured 

on a nominal scale when the observations for the variable represent labels used to 

identify an attribute of each element in the study. For example, variables where the 

data can only be measured on a nominal scale include name, marital status (single, 

married, co-habiting, separated, divorced, widowed etc.), gender (male or female), 

country of residence (Cyprus, Ghana, Brazil etc.) 19 . These labels allow us to categorise 

individual elements into particular groups. The labels can be numerical (e.g., a product 

code), but arithmetic operations such as addition, subtraction, multiplication and 

division do not make much sense on nominal data. 

A variable is being measured on an ordinal scale when the data collected for the 

variable allows for a ranking or ordering to be obtained. A good example of ordinal 

scale data can be found in customer satisfaction surveys where the respondent is 

asked to rate their experience of the organisation (e.g., excellent, good, average, 

poor, very poor) 20 . As with nominal data, ordinal data can be either non-numerical 

or numerical, but again note that arithmetic operations do not make sense. A fivepoint 

Likert rating scale of 1–5 may suggest to you that summing or averaging is 

appropriate but an ordinal scale could quite easily be 23–28. 

A variable is being measured on an interval scale if the differences between 

numerical values are meaningful. To put this another way, can the interval between 

observations be expressed in terms of a fixed unit of measurement? The difference 

between 80 centigrade and 87 centigrade can be expressed through the fixed unit 

18 You may also see time series data referred to as longitudinal or panel data in 

research publications or other methodology texts. 

19 This sort of data could be gathered via a category question, as discussed in 

Section 5. 

20 This sort of data could be gathered via a ranking or rating/scale question, as 

discussed in Section 5. 



of centigrade – in this case 7 centigrade 21 . Such data also possess the properties of 

ordinal data because we can rank or order the observations into warmer (larger) 

and colder (smaller). For variables to be measured on an interval scale requires this 

measurement to be in fixed units, so the data collected will always be numerical. 

With interval data the arithmetic operations are therefore meaningful. As a result, 

data measured using this scale lends itself to more detailed statistical analysis than 

data measured on a nominal or ordinal scale. 

A variable is being measured on a ratio scale if the data has all the properties of 

interval data and the ratio of two observations is also meaningful. Variables such as 

distance, weight and time can be measured using a ratio scale 22 . A requirement of 

the ratio scale is that a zero value exists in the scale. Specifically, the zero value must 

indicate nothing exists for the variable at this point. For example, we would normally 

use a ratio scale to measure income. It is possible to say that employee A on a salary 

of £50 000 is twice as expensive to employ as employee B on a salary of £25 000. In 

addition, Bob, who is currently on work experience with the firm, is free because no 

salary cost is associated with him. The same applies for data measuring numbers of 

children that each element/respondent in the data set has. Since ratio data has all of 

the properties of interval data it is also always numerical and arithmetic operations 

are meaningful. As with interval data, ratio data lends itself to more sophisticated 

statistical analysis than does data measured using a nominal or ordinal scale. 

Thus the amount of information contained in the data varies with the scale of 

measurement. Nominal data contain the least, followed by ordinal, interval and ratio 

data. Furthermore, nominal and ordinal scales can generate both non-numerical 

and numerical data, but interval and ratio scales generate only numerical data. In 

addition we have distinguished between descriptive data and inferential statistics. 

This distinction is important as we proceed. 

Representing Descriptive Statistics 

The purpose of descriptive statistics is to make data manageable, without distorting 

our picture of the subject matter, by summarising the findings. There are many 

ways in which data can be transformed into a more useful form, and there are few 

hard and fast rules about how this should be done. however, there are a number 

of techniques that are tried and tested and we shall focus on these here. There are 

essentially two methods of summarising raw data: graphical and numerical. The 

former provides a general overview of the data without being too precise. The latter 

tends to give a more precise view and can therefore be used as inputs into more 

advanced techniques of statistical analysis. 

21 This sort of data could be gathered via a quantity question, as discussed in 

Section 5. 

22 This sort of data could again be gathered via a quantity question. 



Graphical Representation 

The first step in analysing a data set is usually to present it in a graphical form. 

One of the most common methods used is the frequency distribution diagram or 

histogram. The data in figure 7.1 refers to (imaginary) marks awarded to a group of 

Diploma students, expressed in percentages. We will use this data set to develop our 

histogram. 

25 31 30 41 43 

49 75 65 54 45 

40 50 58 58 79 

68 42 59 53 61 

50 55 55 34 40 

65 38 52 64 67 

Figure 7.1: Foundations of Management assignment marks for August 2009 intake Diploma students 

Figure 7.1 is rather difficult to interpret and appears meaningless. We can establish 

that the highest mark was 79 and the lowest 25, but the overall meaning of the data 

is difficult to see. It would be preferable to have the data in a more presentable form. 

This can be done by dividing the data into a number of classes and constructing a 

frequency table as seen in figure 7.2. 

Class Tally Frequency Relative 

frequency 

20–29 I 1 1/30 

30–39 IIII 4 4/30 

40–49 IIII II 7 7/30 

50–59 IIII IIII 10 10/30 

60–69 IIII I 6 6/30 

70–79 II 2 2/30 

Figure 7.2: Frequencies for the assignment marks data 

Total = 30 



The first column indicates the classes chosen and the second presents a simple method 

of allocating marks to each class. The third then shows summations of these tallies; 

that is, the number of observations falling into each class interval. This is referred to 

as the frequency. The major decision to be made in drawing up a frequency table is 

the width of the class intervals. The narrower the class intervals (and thus the more 

classes) the more the summary table resembles the original data. Therefore, there is 

a trade-off as to how many classes there should be. This is a matter of judgement, 

and depends upon the purpose to which the data is being put and the number of 

observations available. 

The final column of figure 7.2 shows the relative frequency; that is, the proportion 

of the total number of recorded marks which falls into each class. This is calculated 

by taking the class frequency and dividing by the total number of observations. 

however, as you can see in figure 7.3, this information can be presented in the form 

of a histogram. It is much easier and quicker to draw inferences from histograms 

than it is from the raw data or even the frequency table. It is easy to see that the 

observations range from about 20 up to 79, and that most marks are clustered in the 

range 40–69, with few students below 40. We might also suggest that the average is 

in the 50–59 class interval. however, do note that by presenting the data graphically 

we have lost some of the detail, and so we need to return to the raw data to calculate 

the actual average. In other words, as we move from raw data to frequency table to 

histogram we trade completeness and precision for ability to interpret the data easily. 

Frequency 

Assignment Marks 

Figure 7.3: Frequency histogram for the assignment marks data 

It is sometimes useful to present data in a slightly different form to aid interpretation. 

Cumulative frequencies show the number of observations included up to some 

particular point or level. These are calculated by adding successive class frequencies 

together. The cumulative frequency distribution for the assignment results is shown 

in figure 7.4. The final figure of the cumulative frequency must be the same as the 

total number of observations. The cumulative frequency can be useful when we have 

questions such as ‘how many students in this intake failed to achieve a passing grade 

of 50% for Foundations of Management?’ 



Cumulative 

Frequency 

Assignment Marks 

Figure 7.4: Cumulative frequency histogram for the assignment marks data 

It is useful to point out that in histograms the size of the category is represented 

by the area of the bars and not their length. A common error when constructing 

histograms is to overlook this relationship which may produce a distorted view of 

the data. This usually occurs if the data have been grouped into uneven categories. 

For example, if the assignment mark classes began with 0–30, 31–35, 36–41, each 

would represent a different range of marks (30, 5, 6). The corresponding bars in the 

histogram would have to have different widths to maintain the relationship between 

area and category size. 

The histogram is usually used with interval or ratio data. In this case it is possible 

to imagine more measurements between two existing values. For example, with 

the variable of height, a person measuring 170 cm and another person measuring 

171 cm share a continuous scale of measurement (centimetres), but are different by 

1 cm. A third person could measure 170.1 cm and split the difference between the 

two. A fourth person could measure 170.11 cm and split the difference again and 

so on. In contrast, where the values are discontinuous, ‘measurement’ becomes 

impossible and we have to count. In the case of family size for example, we can have 

two adults and two children but we cannot have two adults and 2.3 children. Where 

we are dealing with discrete categories like this we normally use bar charts. Bar charts 

are a type of graph that are used to display and compare the number, frequency or 

other measure (e.g., the mean or average) for different discrete categories of data. 

In figure 7.5 the types of cultural event are the discrete categories of data. 


Percentage of population 

Figure 7.5: Attendance at different types of cultural event in Britain, 1999–2000 23 


Bar charts are one of the most commonly used types of graph because they are 

simple to create and very easy to interpret. They are also a flexible type of chart 

and there are several variations of the standard bar chart, including horizontal bar 

charts 24 or grouped/component/stacked bar charts 25 . The chart is constructed such 

that the lengths of the different bars are proportional to the size of the category they 

represent. One axis represents the different categories and so has no scale. In order 

to emphasise the fact that the categories are discrete, a gap is left between the bars. 

The other axis does have a scale and this indicates the units of measurement. 

In addition to histograms and bar charts, there is a range of other approaches to 

presenting data graphically, including scatter plots, line graphs and pie charts. Advice 

and guidance on the use of these can be found on Blackboard. 

Numerical Representation 

There are several numerical measures that serve to summarise a distribution. The 

most common are measures of location and measures of dispersion. As in the 

graphical presentation of data, there is a choice of techniques available. however nonstandard 

methods are not as readily available in this regard, because the calculations 

performed are often used in further analyses of the data – and these rely upon earlier 

calculations having been performed correctly. 

23 Source: UK National Statistics – http://www.statistics.gov.uk 

24 A horizontal bar chart is one where the bars run from left to right across the 

page as opposed to vertically from base to top. 

25 On a grouped/component/stacked bar chart, a single bar will display different 

categories of data. For example, one bar might show sales of item X for 2007, for 

2008 and for 2009. Each category (sales of item X by year) is therefore displayed in 

relation to the whole (total sales for item X). 



Let’s start with measures of location. These give an idea of where the data is ‘centred’. 

The best known of these is the arithmetic mean, sometimes loosely referred to as 

the average. The arithmetic mean of a data set is defined by taking the sum of the 

observations and then dividing this by the number of observations. Consider the data 

on Diploma assignment marks again, as presented in figures 7.1–7.4 above. The sum 

of the observations would require us to add all of the marks and divide this by the 

number of observations. We could express this as follows: 

25 + … + 79 = 1546 = 51.53 (mean) 

30 30 

But the arithmetic mean is not the only measure of location, and it is not always 

the best measure available. It aims to be representative of the data, yet there are 

circumstances in which it is not. The mean can be significantly affected by outliers 

(a few extreme observations) that pull it away from the more numerous, typical 

observations. In the case of the assignment marks, let’s imagine that four additional 

students failed to submit any work and were therefore recorded as 0 marks. In this 

case, the mean would drop to 45.47 (i.e., 1546 divided by 34). In this example we see 

that the majority of students passed the FoM module, but with the addition of the 

four zero marks the impression of the average mark changes considerably. 

The median is the value within a set of data that divides it into two equal parts. To 

obtain the median, first it is necessary to arrange the data in ascending order. The 

median is then the value that has half of the observations below it and half above 

it. When there is an even number of observations it is necessary to take the mean 

of the middle two observations. Consider the assignment data again. There are 30 

observations. When we put these values into ascending order, we find that the 15th 

value is equal to 52 and the 16th is equal to 53. The median is therefore given by 

(52+53)/2, which equals 52.5. 

The major advantage of the median is that it is hardly affected by outliers. For 

example, if we observed one extra assignment with a mark of only 5, this would have 

a significant effect upon the mean, bringing it down to 50.03. The median, however, 

would now be the 16th observation – that is 52 – hence very little has changed. But 

a problem with the median is that it is not possible to combine two medians to find 

an overall median. If, for example, we know that a factory employs 100 women at a 

median wage of £250 per week and 200 men at a median wage of £320 per week, 

it is possible to calculate the mean wage for all workers without consulting the raw 

data as follows. however, we would have to combine and order all the raw data to 

obtain the new median. 

The final measure of location that is commonly used is the mode. This is defined as 

the observation that occurs with the greatest frequency. Some care is needed when 

determining the mode since it is quite sensitive to how data are grouped. Our raw data 

on assignment marks indicate that the figures 40, 50, 55, 58 and 65 all occur twice, so 

this distribution has no unique mode. Once the data are grouped, however, a single 

class interval with the highest frequency can be found. here the modal group can 

be seen to be the interval 50–59 which has 10 observations. But a different grouping 



of the data might give a different result. Care is therefore needed. We should look 

out for two intervals with similar frequencies, where a slight regrouping might shift 

the mode. We should also look out for unequal class widths, since combining classes 

is bound to increase the associated frequency. Indeed the mode could be shifted 

anywhere by combining sufficient class intervals! The best advice is to calculate the 

frequencies as if the class widths were equal and then choose the modal group. 

Moreover, individual data can be distributed randomly in the data set, appear 

predominantly towards the bottom of the value range, or predominantly towards the 

top. This influences the relationship between mean, median and mode. Consider the 

frequency distribution shown in figure 7.6, which is drawn as a smooth, continuous 

distribution curve for convenience. 

Figure 7.6: A positively -skewed distribution 

It is unimodal (has a single mode) and skewed to the right, i.e., the longer tail is to 

the right. In this case the mean (also referred to using the symbol μ) is always greater 

than the median (M e ), which is always greater than the mode (M o ). If the distribution 

were skewed to the left, then the order would be reversed. A symmetric distribution 

would have the mean, mode and median in the same place. 

Now let’s move to consider measures of dispersion. So far, we have distinguished 

data distributions on the basis of their ‘central values’; for example, their means. Two 

distributions with the same mean could still be very different, however, as shown in 

figure 7.7. 



Figure 7.7: A comparison of two different distributions with identical means 

The distribution for data set X 2 is clearly more spread out (dispersed) than the 

distribution for X 1 . If these were two sets of assignment marks, assignment 2 

would have far more grades at the higher and lower ends of the grading scale than 

assignment 1, even though both data sets share the same mean. 

Clearly it would be useful to have a numerical measure that summarised the different 

degrees of dispersion of these distributions. At first sight it might seem as if such 

a measure of dispersion could be constructed by considering the mean deviation 

around the mean – the mean of the differences in value of all the data from the mean. 

Consider the following two sets of data: 

(i) 2 4 6 8 10 

(ii) 4 5 6 7 8 

In each set the mean and median are 6. however, the data for (i) differ from (ii) in that 

they are more dispersed. We can construct the mean deviation by subtracting the 

arithmetic mean from each individual value. For example, with data set (i), the mean 

= 6 and the first value is 2. Therefore 2 – 6 = –4, and so on. But does the sum of these 

differences give the total dispersion of the sample? This is the result of summing the 

differences between each value and the mean for data set (i). That is: 

(–4) + (–2) + (0) + (2) + (4) = 0 

The problem with this measure is that the positive and negative deviations around the 

mean will cancel each other out. To remedy this problem the deviations are squared 

(multiplied by themselves – e.g., 4 squared is 4 x 4 = 16) before being added together: 

(–4) 2 + (–2) 2 + (0) 2 + (2) 2 + (4) 2 =16 + 4 + 0 + 4 + 16 = 40 



The sum then needs to be scaled in some way to take account of the number of 

observations (5): 

40/5 = 8 

When this is done, the measure that results is called the variance. The question then 

arises as to what units the variance is measured in. Since the deviations were squared 

in the course of the calculations, the answer must be that these are now in squared 

units. This is a somewhat difficult concept to grapple with, so we return to ordinary 

units by taking the square root of the variance, to obtain the standard deviation. 

The square root is the value which, multiplied by itself, produces the value you began 

with, and can be arrived at using even the most basic calculator. Just enter the value 

then press the √ key. With a variance of 8, the square root would be 2.82. 

The standard deviation is explained succinctly by Rowntree (1981:53–54) as follows: 

“If there were no dispersion at all in a distribution, all the observed values 

would be the same. The mean would also be the same as this repeated 

value. No observed value would deviate or differ from the mean. But, 

with dispersion, the observed values do deviate from the mean, some 

by a lot, some by only a little. Quoting the standard deviation of a 

distribution is a way of indicating a kind of ‘average’ amount by which 

all of the values deviate from the mean. The greater the dispersion, the 

bigger the deviations and the bigger the standard (average) deviation.” 

Two measures have thus been presented that can be used to summarise a distribution 

or to compare different distributions; the mean and the standard deviation. 

The standard deviation can also be used to compare individual observations from two 

distributions. Suppose there are two groups of students studying the same module. 

The module assignment for each group is set and marked by different members of 

academic staff, but there is only one prize to award to the best student overall. Should 

it be awarded to the best student in Group 1 or Group 2? It would be unfair simply to 

take the student with the highest mark, since one assignment could be more difficult 

than the other, the marking more severe in one case, or different marking scales may 

have been used. 

One solution would be to give the prize to the student who has done best relative to 

his or her own group. This could be measured by the number of standard deviations 

each student’s mark is above or below the mean of their group. Only the marks of 

the best student in each group would have to be used, so the calculation is fairly 

straightforward: 

69 – 60 = 1.5 55 – 50 = 2.0 

6 2.5 

(Group 1) (Group 2) 



Thus the prize should be awarded to the student from Group 2, since to obtain a score 

of 2.0 standard deviations above the mean is probably a greater achievement than 

the 1.5 achieved by the best student in Group 1. The assumption of this approach, 

however, is that the two groups are broadly similar to begin with. The variance is 

the most widely used measure of dispersion because, like the mean, its statistical 

properties can be established. But there are other measures, which also have their 

uses. The range is simply the difference between the largest and the smallest 

observations. It is easy to calculate once the data has been ranked, but it has little 

else in its favour. It makes use of only a small amount of information in the sample 

and is obviously very sensitive to outliers. 

having discussed ways of representing descriptive statistics, our last topic of discussion 

in this section of the module book is statistical inference – i.e., using samples to 

make inferences about wider populations. 

Statistical Inference 

The aim of a statistical investigation will usually be to find out the characteristics of 

the relevant population, but on many occasions (as we saw in Section 5) it will be 

impractical or impossible to collect data on every element of this population. Thus, 

once a sample has been taken, the next step in a statistical investigation is to work out 

how accurate that sample is by evaluating its various characteristics. Characteristics 

of a population in this instance are referred to as parameters, while characteristics 

of the sample are called statistics. 

There are two ways in which sample data can be used to make inferences about the 

characteristics of a population. The first involves the construction of a confidence 

interval using the sample data. For example, if a sample of 1000 units from a 

production process is taken and 5% are found to be defective, what can be concluded 

about the range into which the population percentage falls? In this case, because 

sample data are being used, there is no guarantee that 5% of the items overall will 

be defective – but it would be very unlikely that 25% were defective. 

The second is hypothesis testing (as previously discussed in Section 4) – the 

specification of a hypothesis and its testing using the sample data. here, we might 

start with the hypothesis that only 4% of the aforementioned items are in fact 

defective. Then the sample information, which indicates that 50 out of 1000 were 

found to be defective, would be used to determine whether this sample could have 

come from a population in which only 4% are defective (the hypothesis). here the 

difference between the two would be due to chance factors, particularly in view of 

the fact that only a sample is being used. 



In solving these sorts of problems it is assumed that the sample being used is taken 

randomly. A random (/probability) sample, as we saw in Section 5, is a sample selected 

so that every member of the population has an equal chance of being selected; and, 

as we have also seen, there are various ways of doing this. This assumption is crucial 

for appropriate statistical inference. 

Clearly, there is no guarantee that the statistic will be identical to the parameter, but 

one would expect there to be some relationship between the two. This notion can be 

expressed in a very general form: 

Sample statistic = Population parameter ± Sampling error 

If we rearrange this, we see that the population parameter will be related to the 

sample statistic as follows: 

Population parameter = Sample statistic ± Sampling error 

This formulation shows that, in order to establish the relationship between parameter 

and statistic, it is necessary to identify the sampling error, as also discussed in Section 

5. In other words, we need to work out the sampling distribution of the sample 

statistics. This can be done by addressing the situation where a number of samples 

are taken from a population and looking at the distribution of the sample statistics 

that result. Furthermore, as the sample size increases and begins to approach the 

total number of observations within the population, one would expect the statistic 

to tend towards the parameter; that is, the sampling error gets smaller. 

Remember that the calculation of a statistic is made on the basis of sample data. 

The value of the sample mean, for example, will vary from sample to sample and 

can be regarded as a random variable. The reason for this is that each set of sample 

data will (or at least should) consist of different members of the population, and 

therefore any calculations based upon it will also vary. If we generate samples from a 

known population we will generate a set of sample means, a set of sample standard 

deviations, and so on. These sets of data can be treated in the same way as any others, 

and we can obtain their means and variances as well as identifying the distribution 

involved. What we are interested in, then, is the relationship between the sampling 

distribution and the population distribution. 

Although sampling distributions can be generated empirically, as outlined above, 

it is often better to derive mathematical formulae for sampling distributions. This 

approach enables us to apply the results to many sampling problems. Thus every 

statistic can be treated as a random variable and it will have a particular distribution, 

which is called the sampling distribution of the statistic. Such distributions have 

long been recognised within the statistics literature with examples including the 

normal distribution, the student t distribution and the F distribution. These 

established distributions can be used to form the link between population parameter 

and sample statistic since they accurately depict the nature of the sampling error. 



In the case of the construction of a confidence interval, values taken from the 

appropriate distribution would be used to formulate a range into which the 

population parameter is likely to fall. Note the word ‘likely’ since this indicates that 

there is always a chance of being wrong. This chance or probability is sometimes 

referred to as a p value or significance level. Values taken from the appropriate 

distribution allow us to control for this chance of error, but it can never be completely 

eliminated. In the area of statistical analysis researchers typically specify either a 5% 

or 1% chance, hence a 5% or 1% significance level. Again we saw this in Section 5. 

To illustrate this, consider the example from earlier. A sample of 1000 units from a 

production process is taken and 5% are found to be defective. In this situation values 

would be taken from the normal distribution according to its standard table of 

probabilities. These values, together with other information on variability and sample 

size, enable the identification of the sampling error and hence the construction of 

a confidence interval. In the case of a 5% chance of error (i.e., the construction of a 

95% confidence interval), the range would be 3.65% through to 6.35%, indicating 

that we can be ‘95% confident’ that the overall percentage of defective units coming 

off the production line lies between 3.65% and 6.35%. If instead the researcher chose 

a 1% chance of error (i.e., the construction of a 99% confidence interval), the range 

would be 3.22% through to 6.77%, indicating that we can be ‘99% confident’ that 

the overall percentage of defective units coming off the production line lies between 

3.22% and 6.77%. Note that in this situation the interval has widened, showing that 

there is a trade-off between confidence and precision. 

With hypothesis or significance testing the 5% and 1% figures indicate the 

chance or probability of drawing the wrong conclusion from the test, hence a 5% 

or 1% significance test. Values would again be taken from the standard table of 

probabilities. These values and information on variability and sample size would be 

used to calculate a test statistic, and the result of this calculation would lead us to 

accept or reject the hypothesis that only 4% of the items are defective. Given the 

information above (5% defect rate in the sample of 1000), the hypothesis of an 

overall defect rate would be accepted. Note that this is consistent with the confidence 

interval calculation as 4% falls within both confidence intervals. 

In this section we have looked at some of the basic principles that ground some of the 

assumptions surrounding quantitative data analysis. We have tried to avoid going into 

detail with complex mathematical equations. Rather our aim was to encourage you 

to see the underlying rationale behind the principles. We identified the relationship 

between the type of data you might be looking at and the possible techniques that 

can be employed to analyse that data. We have also discussed the ways in which 

we can represent data, both graphically and numerically. Lastly, we explored the 

relationship between samples and populations. Using these relationships between 

different types of data, you can begin to convert these practices into quantitative 

analysis in your own work. 


Summary 


1. Descriptive statistics involve numerical descriptions of observations; inferential 

statistics involve making estimates or predictions based on these observations. 

2. Data, whether quantitative or qualitative, can be classified as cross-sectional 

or time series. 

3. Quantitative data can be measured on a nominal, ordinal, interval or ratio 

scale. 

4. Descriptive statistics can be represented using graphs or numbers. 

5. Inferential statistics involves either the construction of confidence intervals or 

hypothesis testing. 


� Tasks 









References 





University Press chapter 14 



Rowntree, D. (1981) Statistics Without Tears: A Primer for Non-Mathematicians 

London: Penguin 










section 8 

Analysing Qualitative Data




SECTION 8 

Analysing Qualitative Data 



• identify the possible errors in analysing data of any kind 

• discuss the merits and limitations of different strategies for 

analysing qualitative data 

• identify the pros and cons of qualitative data analysis software 

• consider the different ways to present analysis of qualitative data, 

and whether or not you need to produce a full transcript. 

Introduction 

This section deals with some of the key issues around the analysis of qualitative data. 

Qualitative data, as we know, are collected as words, through more flexible research 

methods such as semi-/unstructured interviews or non-structured observation. 

Analysis is then typically conducted by classification of these data into categories 

or themes. Qualitative data analysis is usually written up in the form of a textual 

narrative interspersed with direct quotations from the respondents/participants. 

As we have also established elsewhere, qualitative data are often understood as 

providing a richer and more complex or in-depth account of an empirical site. As 

Robson (2002:454) suggests, “Narratives, accounts and other collections of words are 

variously described as ‘rich’, ‘full’ and ‘real’, and contrasted with the thin abstractions 

of number.” however, this is not always the case, especially if content analysis is used 

to analyse the resulting data or the data collected are not fit for purpose. 

Data Analysis in General: A Brief Overview 

To ensure that you collect the right type of data for your project, as well as ensuring it 

will actually enable you to explore the areas you are interested in, you should consider 

how you will analyse your data very early on. Indeed data analysis should be a factor in 

several other methodological decisions. In other words, when selecting your method, 

choosing your sample, designing your schedule (as suggested in Section 5), choosing 



your channel, piloting your schedule and doing the data collection itself, you should 

think about the data you need to answer your research questions, and how you will 

do the analysis. If you haven’t given these issues some serious thought beforehand, 

you may find that it is very difficult to interpret the data you actually end up with. 

Analysing any kind of data, furthermore, is about data reduction – i.e., making sense 

of a data set, reducing it to manageable and meaningful proportions. But there are 

fewer conventions or established practice when it comes to analysing qualitative 

data as compared to its quantitative counterpart, where you apply specific statistical 

techniques depending on what you want to achieve (as we have seen in Section 

7). So to some degree rather more consideration and judgement is required in the 

selection of an analytical strategy for qualitative data. Furthermore, with qualitative 

data in particular, try to analyse them as soon as you can after collection. This way the 

research encounters are fresh in your memory and you remember them more clearly, 

especially as regards any problems ‘in the field’ which may have affected the data in 

particular ways. 

Data Analysis and Perceptual Errors 

Before we move on to examine particular approaches to the analysis of qualitative 

data, it is also worth pointing out that, when analysing data of whatever kind – 

including data from research projects – human beings have certain limitations due 

to the way our cognitive processes work. As a result we may make certain common 

errors in such analysis. These are effectively perceptual errors. These errors are worth 

bearing in mind when analysing your data, because you should try to avoid them. 

• Data overload: our brains can only cope with a certain amount of information 

at once in terms of taking it in and trying to understand it. You may therefore 

have to do your data analysis in stages to accommodate this. It also means 

that the more data you have the more time you will need to analyse them – 

and there are feasibility issues here, as noted in earlier discussion in this book. 

• First impressions/last impressions: because of how our perceptual processes 

operate, we may form a judgement very early on – i.e., a first impression – of 

what is going on in a particular data set. This is problematic because it means 

we may be tempted to ignore contradictory data or make too much of data 

that support these first impressions as we continue our analysis. Try to keep 

an open mind throughout and not to allow early judgements to sway your 

conclusions. You should in any case, especially when working with qualitative 

data, revisit the data set several times to ensure that what you thought you 

had identified at first is actually there. Equally though, we might be swayed 

by information which presents itself towards the end of the data set (last 

impressions), and so the above also applies in reverse. 

• Assuming internal consistency: relatedly, we may assume (usually wrongly, 

especially with qualitative data) that data are consistent throughout. So when 

we come across information which conflicts with what we have seen earlier in 

the data set, we might ignore it. But human beings are unique and complex 



creatures and we are also not necessarily consistent as individuals. Indeed we 

may say one thing at one point (and mean it) and then say the reverse later 

in an interview (and also mean it). Don’t therefore ignore any complexities 

and contradictions. Instead, try and understand why they might be present in 

your data. 

• Being persuaded by information availability: we can often be persuaded 

by information which has been easy to get hold of, and neglect to access 

other data which might require more effort but tell a different and equally 

important story. For example, it may be relatively straightforward to talk to 

managers, who are likely to offer one story about the success of a particular 

organisational change initiative, but more difficult to access rank-and-file 

employees, whose perceptions might be rather different. 

• Uneven validity: some sources of data can be regarded as more valid or 

truthful/accurate than others, and you should pay attention to this in your 

data analysis. An example is that secondary data from media reports is often 

highly coloured by the political bias of the source. Another example is that 

particular respondents may have their own agendas which are reflected in 

the data you gather. We could, as implied above, assume that managers will 

generally want to paint their organisations in a positive light. Another case is 

a paper that a member of ULSM faculty reviewed recently. The author of the 

paper had interviewed several people in an organisation to assess whether 

the manager in charge of the group (a homosexual man) tended to favour 

other homosexual men in his decision making. The heterosexual respondents 

all said that he did. The single homosexual respondent suggested this was 

not the case. The researcher assumed that the heterosexual respondents were 

more trustworthy than the homosexual respondent in this instance, because 

the latter had probably benefited from the manager’s favouritism and not 

necessarily recognised it as favouritism. But this isn’t necessarily true. Perhaps 

the man in question had in fact done well on his own merits and was resented 

by his heterosexual colleagues, who then developed a rather different version 

of events to protect their self-esteem. You need to think carefully about issues 

like this – i.e., what might be affecting the quality or veracity of the data. 

• Conflating co-occurrence/correlation with causation: a common error 

made when analysing quantitative data in particular, but also qualitative 

data, is to confuse co-occurrence or correlation with causation. If one variable 

changes in the same way and at the same time as another, this does not 

automatically mean that one causes the other. Two examples should illustrate 

this point. The first is the data which suggest that, as the use of flushing 

toilets increases, so do levels of coronary heart disease. So does this mean 

that flushing toilets cause heart disease? In fact what these data tell us 

is that, as standards of living rise and so more people can afford flushing 

toilets, coronary heart disease (ChD) also rises because more of us are doing 

better paid, but also less active, more desk-bound, occupations, as well as 

eating richer diets (which are more expensive but often also less healthy). 

Less activity and less healthy diets lead to higher rates of ChD. So there is a 



co-occurrence/correlation here, in the sense that a rise in flushing toilets and 

a rise in ChD occur together; but the relationship is not causal. The second 

is the connection which was found between guardsmen fainting whilst on 

parade and the softness of the tar on the parade ground. One explanation 

of this is that the unpleasant fumes from the melting tar caused the men to 

faint – but a more likely scenario is that hot weather causes both fainting 

and melting tar (cited in Robson 2002:484–485). Again here we have a cooccurrence 

or a correlation – not a causal relationship. 

• Researcher inconsistency: although as we have said above we would 

recommend that you revisit your qualitative data several times to check your 

interpretations, it is also true that we as researchers are inconsistent and so 

we may well change our minds when we see a data set again. This however 

may simply indicate that the data set is in fact quite complex and rich. Also, as 

long as you can justify the interpretation that you include in the final project, 

even if there are others, this is not a problem. 

What we will now outline are a number of strategies which you may wish to consider 

if you think you will need to analyse qualitative data of some kind. One of the first 

issues you need to consider in this regard is your research purpose. We have already 

argued in Sections 1 and 5 that research is usually more than just a description of 

what is going on in a particular area. however, you may need to do some descriptive 

analysis to provide background or context – e.g., to paint a picture in your assessors’ 

minds of the organisation/s where you did your data gathering. Explanatory/ deductive 

research as we also know wouldn’t be typical of a qualitative project, but this does 

not rule it out entirely as a possibility. We will therefore say something briefly about 

this sort of approach, outlining two basic strategies for such analysis, offering an 

example of deductive qualitative work and outlining its pros and cons. Nonetheless 

we will spend more time on exploratory/inductive research as this is more ‘classically’ 

qualitative. 

Deductive Analysis of Qualitative Data 

Pattern Matching 

Our first strategy here is pattern matching, a verificationist strategy where you literally 

search the data to see how far their patterns verify or match themes suggested by the 

hypothesis you are testing. Then you would accept, modify or reject the hypothesis 

on this basis. This would usually imply looking either for changes in dependent 

variables (specific outcomes – for example, the results of a change management 

process) or independent variables – for example, trying to identify the causes of 

known outcomes like a decline in productivity or an increasing number of workplace 

accidents. 



Negative Case Analysis 

Negative case analysis involves searching through the data for negative cases which 

disprove the starting hypothesis and mean it has to be modified to better fit the 

data in this circumstance. The example that is usually given is the hypothesis that ‘All 

swans are white’. here the researcher would deliberately look for non-white swans 

in the data. Negative case analysis is based on Karl Popper’s falsificationism. Simply 

put, Popper argued that we can only ever accept a hypothesis on a provisional basis, 

because we never have all the data available to be absolutely sure that it is true 

in every single pertinent case. however, Popper argues that we can be sure about 

rejecting a hypothesis where we find contradictory data; so he argues that we should 

always look for such data. In this example we would probably seek to modify the 

hypothesis after finding some non-white swans (e.g., ‘Some swans are white and 

some are black’) and test it again. 

Popper argues that, if we don’t follow a falsificationist strategy when seeking to 

evaluate hypotheses, then we tend to attend only to data which support our existing 

beliefs and so confirm the starting hypothesis. This, he suggests, means that we 

find it difficult to develop new and perhaps improved ideas about human behaviour. 

Any hypothesis which survives a negative case analysis of this kind would therefore 

be corroborated (because the data gathered in this research project support it) 

but not confirmed (because the likelihood, according to Popper, is that there are 

data elsewhere in the world which would lead to its modification or even outright 

rejection). 

An Example of Deductivism in Qualitative Management Research 

One example of an explanatory purpose or deductivism/theory testing in qualitative 

research is Scott’s ethnographic study of hRM in British organisations. he wanted to 

test the hypothesis that “British management was operating according to a ‘new’ 

model of industrial relations, based on a unitaristic view of organizational life, where 

workers and managers are seen to have a similar interest in the success of the firm” 

(Bryman and Bell 2007:628). In other words, Scott wished to discover whether, given 

the widespread introduction of hRM as an alternative approach to managing staff 26 , 

British employees and managers did in fact have similar interests (the unitarist 

philosophy which hRM is based on), or whether differences persisted in the form 

of ‘them and us’ type cultures. Scott’s findings indicated that any movement away 

from ‘them and us’ pluralism to hRM unitarism was incomplete and only found in 

organisational ‘pockets’ here and there. So his starting hypothesis was not confirmed. 

Pros and Cons of Deductivism in the Analysis of Qualitative Data 

The strength of doing deductive analysis of qualitative data (which of course would 

have been gathered to test a specific hypothesis) is that it gives you a clear path 

through the data. In other words, the hypothesis tells you what to look for in the 

data. But, as Bryman points out, 

26 Refer back to Foundations of Management for discussion of hRM. 



“The prior specification of a theory tends to be disfavoured [by qualitative 

researchers] because of the possibility of introducing a premature 

closure on the issues to be investigated, as well as the possibility of 

the theoretical constructs departing excessively from the views of 

participants in a social setting.” (cited in Saunders et al. 2009:489) 

In other words, it has been argued that deductive work in this type of research may 

well lead to ‘premature closure’ of emergent issues and a neglect of complexities 

and nuances in the data because the hypothesis limits what the researcher looks at. 

This can mean that some parts of the data are ignored altogether – whether they 

support the hypothesis or not – since they appear not to have anything to do with 

the variables which the hypothesis identifies as significant. Testing a hypothesis may 

therefore make it difficult to see qualitative data holistically or ‘in the round’ – to 

excavate all its richness, which is arguably its real strength when collected properly, 

as compared to quantitative data. 

And so we move to inductive analysis of qualitative data, which is a more common 

approach. 

Inductive Analysis of Qualitative Data 

Grounded Theory, and a More Realistic Alternative 

Grounded theory is perhaps the most extreme form of inductivism when analysing 

qualitative data. It was originally developed by Glaser and Strauss (1969) in the 

late 1960s but they subsequently had an intellectual disagreement and have since 

published rather different accounts of what it means to conduct grounded theory 

research (Glaser 1978, 1992; Strauss and Corbin 1998). In many ways grounded 

theory can be understood as a methodology in itself – it is arguably more than just 

an approach to data analysis. What follows is an attempt to summarise some of the 

key aspects of this complex and controversial approach to qualitative research and 

analysis. 

First, and because of the emphasis on inductivism here, the idea is that the eventual 

theory (the way in which the researcher finally understands what is going on in the 

data) emerges from/is grounded in the data alone. It has not been informed in any 

way by pre-existing hypotheses. This means, grounded theory proponents argue, 

that the researcher doesn’t force the data to fit the theory (a potential problem 

of deductivism according to Popper, as we have already seen) but instead develops 

a theory which fits the data. Therefore strictly speaking grounded theory requires 

that the first step of the process is to design your schedule and collect and analyse 

your data. This happens before any consultation of the relevant subject-specific 

literature on organisational structure, organisational culture, relationship marketing, 

charismatic leadership, stress, productivity, employee motivation, the supply chain, 

auditing, whatever. An MBA student we once taught also described the need to ‘have 

a good wash’ before data collection using grounded theory begins, to get rid of as 



much pre-judgement or bias as possible, and to go in with an open mind. Grounded 

theory methods typically involve unstructured interviewing 27 or some form of nonstructured 

observation. They usually also gather qualitative data. 

however, although you may wish to read more about grounded theory for your 

own interest or even for your PMP, it has a number of drawbacks. For instance, it 

takes a very long time to do properly. In fact as already established it is more of a 

methodology in and of itself than purely an approach to analysing qualitative data. 

Second, it is crucial to be precise in terms of following all the recommended steps 

to the letter. So, rather than describing grounded theory in detail, we are going to 

discuss what we feel is a more realistic alternative. We would suggest that this is a 

more common way to analyse qualitative data. 

This approach is a variation on grounded theory. So it takes its inspiration from 

grounded theory but does not follow this methodology in its entirety because of the 

above deficiencies. Overall it is partly inductive but more structured/guided from the 

outset by the subject-specific literature and less reliant on repetition of steps than 

grounded theory. 

It proceeds as follows: 

• Consult subject-specific literature and inform yourself about the organisation/s 

or context/s where you are gathering your data. As discussed in Section 5, use 

these sources of information to develop your research schedule. 

• Pilot if possible, then amend your schedule accordingly and gather your data. 

• Familiarise yourself with the data, which involves you immersing yourself in 

the data; reading through them carefully and beginning to note emergent 

themes. 

• Coding: “A code is a symbol applied to a section of text to classify or categorize 

it” (Robson 2002:477). A different code should be applied every time the 

emphasis or focus of the data seems to change – which could even be several 

times in one sentence. Examples of codes might be “‘requesting information’ 

27 The MBA student referred to above actually had only one question in her 

unstructured interview schedule. her research set out to assess whether a particular 

change programme involving the devolution of the human Resource function from 

a centralised department down to separate directorates and a general shift towards 

an hRM approach away from a personnel management approach in a UK NhS Trust 

had been successful. her sole question, to those who were affected by the change, 

was ‘how was it for you?’ her findings suggested that the change had been very 

‘top down’ – i.e., imposed by senior management – so individuals had no choice but 

to re-educate themselves about the benefits of devolution and of the shift towards 

hRM. Overall then the change had been received as positive, despite the fact that 

the way it was managed contradicts most if not all the available literature on change 

management (Smart 1994)! 



or ‘expressing doubt’” (Robson 2002:461). The basic question here is ‘What 

is this piece of data an example of?’ You should of course bear your literature 

review and your overall research questions in mind when identifying codes – 

what do these suggest you might expect to find in your data? 

• Also be open to any emergent codes and contradictions. Pay careful 

attention to anything new which you weren’t expecting as well as being alert 

to contradictions and complexities in the data. For example, a co-author of 

this book, when conducting her PhD research on sexual relationships at work, 

kept noticing ‘space’ as a theme in the data she had collected – i.e., the 

idea that the physical design of office space affects relationships between 

colleagues. This was not something she had expected to find, but it was too 

strong a theme to ignore. 

• Reread your data, reread your analysis – do the two still seem to fit together? 

• Integrate literature and data: refer back to the subject-specific literature to 

see the extent to which the analysis seems to conform to, alter or entirely 

depart from the conclusions drawn by other authors on the same subject. 

• Return to respondents to check validity if you can: respondent validation of 

the analysis of qualitative data is always a good idea if possible – you can, for 

example, send a draft by e-mail for comments. 

• Produce the final version of the analysis. 

But qualitative researchers may also employ a technique called content analysis which 

as we have suggested already literally counts or quantifies qualitative data and thus 

turns it into quantitative data. Content analysis can be used in either deductive or 

inductive research. We offer a brief overview of what it is and how it works next. 

Content Analysis 

As an example, hodson (cited in Bryman and Bell 2007:319, 636) undertook content 

analysis of 86 workplace ethnographies 28 . here he counted how many organisations 

fell into the various categories described (‘craft’, ‘direct supervision’, ‘assembly line’, 

‘bureaucratic’ and ‘worker participation’) and treated ‘organisational category’ as the 

independent variable. hodson then sought to see how this variable affected various 

dependent variables, whilst also attending to potentially mediating variables – 

which intervene between independent and dependent variables to produce different 

effects. Two examples here are job satisfaction and level of worker autonomy. 

28 In other words, published accounts of research in workplaces that had been 

carried out using participant observation (refer back to Section 5). 



hodson concludes from his study that academic pessimism about whether employee 

participation schemes actually benefit the employee or are genuine in their intentions 

might be misplaced. he also argues that his approach has potential to become the 

basis for future research projects as ethnographic accounts of organisations increase 

in number. hodson acknowledges that content analysis ignores contextual factors 

because it singles variables out from the data to explore relationships between 

them. however he suggests that, unusually for qualitative research, it allows 

coverage of a range of different types of organisations as well as hypothesis testing/ 

deductivism. Furthermore, the data he analysed are more detailed and rich than 

standard quantitative data. 

To undertake content analysis, consider the following steps: 

• Recording unit: What level of analysis will you apply to the data? Are you 

going to attend to each word, each phrase, each sentence or each paragraph? 

• Categories: These are the themes or variables you are looking for. Some will 

come from the subject-specific literature, especially if you are seeking to test 

a specific hypothesis. Some will be themes suggested by the data themselves. 

The list of categories needs to be relevant to the research questions you want 

to answer. It should also be as exhaustive or comprehensive as possible in this 

regard, contain mutually exclusive categories and be clear and easy to use. 

Indeed this stage of content analysis is similar to the design of structured 

observation schedules, as discussed in Section 5. 

• Test and revise: Pilot the list of categories you have chosen against the data 

to check that it will fulfil the criteria outlined above, and revise if necessary. 

• Analyse the data themselves. 


Example 8.1: An example of a content analysis schedule 


This schedule could be used to quantify reasons given by eight 

respondents for the introduction of an incentive scheme in a given 

organisation. It has been adapted from the first edition of Easterby- 

Smith et al. (1991:107).The content analysis coding sheet might look 

something like this: 

To increase profits 

To increase 

productivity 

To increase 

efficiency 

To increase labour 

flexibility 

To increase earnings 

To reduce 

production costs 

To reduce 

absenteeism 

To reduce wastage/ 

rejects 

etc … 

1 2 3 4 5 6 7 8 Total 

You would then count the number of times each respondent mentioned 

each reason and total them up to arrive at a set of figures for further 

analysis. 

Content analysis can be seen as beneficial because it effectively ‘cuts the data up’ into 

themes/categories/variables and allows you to count the frequency of occurrence 

of each. It can be therefore seen as offering more objectivity compared to other 

methods of analysis of qualitative data. It is also easier to use than the inductive 

method we described above. But its drawbacks include the following: 



1. Saunders et al. (2009:497) note that “there is clearly only limited purpose 

in collecting qualitative data if you intend to ignore the nature and value 

of these data by reducing most of them to a simplified form”. So content 

analysis could be a good strategy to use in initially getting to grips with a 

qualitative data set. But it does tend to downplay the complexity and nuances 

of qualitative data, as well as ignoring its context, as already established 

above. 

2. You should also try not to be seduced by the fact that content analysis 

generates numbers, and so may feel rather more ‘scientific’. As we have tried 

to establish throughout this book, there is no one best way to do research. 

Instead your choices should be guided in the main by what works best for 

your project in your circumstances. In any case, content analysis still involves 

researcher interpretations about what respondents mean and how they use 

language, and thus how to categorise the relevant data – so there is still a fair 

amount of subjectivity involved. 

3. Content analysis can be argued to neglect respondent interpretations in the 

sense that it may tell you what they feel, say or do but not why they feel, 

say or do these things. For example, Barley et al. (cited in Bryman and Bell 

2007:304, 321) wanted to research the ways in which management academics 

and managers influenced each other. They chose to do content analysis on 

192 articles on organisational culture 29 , some written by academics, some 

by managers, and published between 1975 and 1984. Their findings tell us 

that the academic papers gradually took on board managers’ interpretations 

of culture over time, but that the reverse was not true. however, what this 

analysis fails to tell us is why this might have been the case. 

Moving on, we now consider the extent to which software packages might assist in 

the analysis of qualitative data, given their popularity in the analysis of its quantitative 

counterpart. 

Software Packages and Analysing Qualitative Data 

WARNING! Any software package is only as good as the data (whether quantitative 

or qualitative) you have collected. Qualitative packages also don’t do the job for 

you to the extent that SPSS or spreadsheets 30 might with quantitative data – they 

certainly cannot substitute for your interpretive skills. 

29 Refer back to Foundations of Management for discussion of organisational 

culture. 

30 SPSS stands for Statistical Package for the Social Sciences and is a 

sophisticated software package for quantitative data analysis. however, most of the 

analysis you will probably need to do for any PMP involving quantitative data can be 

accomplished using an Microsoft Excel spreadsheet, or a decent calculator 



Some types or functions of qualitative data analysis software packages (Weitzman 

and Miles, cited in Robson 2002:462–463) include: 

• text retrievers 

These can find certain words or strings of words or characters (i.e., letters 

and numbers). Some text retrievers count and display words in their context 

as well, so you can see in general where these tend to appear in the data, as 

well as how frequently they appear. however, the search function in a wordprocessing 

package like Microsoft Word will also find words/strings of words 

or characters for you. 

• code and retrieve programs (e.g., Ethnograph) 

These help the researcher to divide the text into segments/recording units and 

to attach specific codes to these segments. They will then find and display all 

the text which is coded in a particular way. Again though there are ways to 

do this using Word – e.g., by highlighting some text in italics as belonging to 

a specific code and then instructing the package only to find italicised text 

using the search function. 

• code-based theory programs (e.g., NVivo) 

These have the same functions as code and retrieve programs but also help 

to support theory building because they can allow a researcher to connect 

codes together as well as developing and testing emerging propositions 

against the data. These programs would be particularly useful for grounded 

theory. 

Things to bear in mind if you want to use such a package are, first, that your judgement 

and interpretation still matter, as suggested above. Further, choose carefully according 

to criteria like its cost, whether any preparation of the data is required before the 

package will accept them, your own style of analysis and what the package actually 

does. You should also remember that you will need to be proficient in operating the 

package before you do the actual data reduction. Most software companies will 

allow you to trial a package before you buy it. Third, remember that using a software 

package to analyse qualitative data might lead to an unwitting emphasis on ‘word 

crunching’ and focusing on quantities of data in themes and categories as opposed 

to understanding the meaning, complexity and context of the data. Manual analysis, 

in short, is probably easier in the long run and allows you to get much closer to your 

qualitative data in all its complex, variegated and ambiguous glory. 

As an example, halme (2002) – whose research we discussed in Section 2 – used a 

qualitative data analysis package in her grounded theory research. She said this made 

it easier for her to follow all the various stages of coding and recoding this required 

because the software speeded this process up. But she does emphasise that the 

package did not do the thinking for her and that she effectively used it in the same 



way as we might use a series of index cards to record codes, or coloured markers to 

highlight coded data on a hard copy transcript. 

Presenting Qualitative Data 

The presentation of qualitative data analysis is usually not just descriptive, as we have 

already argued. What follows are therefore some pointers to make your presentation 

meaningful. Some of these techniques are also useful for analysing data. 

• narrative and quotes 

As suggested above, this is the standard approach which can be found in 

the majority of publications based on qualitative research. The researcher 

here intersperses direct quotations from the data with indirect quotations 

summarising respondents’ feelings, behaviour and beliefs 31 . Theoretical 

conceptualisation – i.e., what these data say about the relevant literature 

and your research questions – may be presented at the same time or as a 

separate, subsequent chapter. however, in a 6000 word project, we strongly 

recommend that you present and analyse the data in the same chapter to cut 

down on length. 

• matrices 

here the researcher would use a spreadsheet where each respondent/ 

group of respondents is presented on an individual row, and the columns 

would indicate particular themes or issues in the data. Each box would then 

summarise what each respondent/group had to say about each issue, perhaps 

with an indicative direct quotation or two. 

• diagrams and charts 

These could for example depict lines of travel around a workplace derived from 

observation data, an organisation structure or flow charts of work processes 

derived from interview or observation data. It might also be interesting to 

compare any such data with ‘official’ documents where they exist – e.g., to 

the organisation structure as it appears in the Annual Report, for example. 

• causal networks 

These would trace independent and dependent variables and/or correlations 

between variables to show where the relationships exist and what the 

outcomes are, using pictures of these causal networks. 

31 Refer back to the guidance on referencing in Foundations of Management or 

your programme handbook for a reminder of what the difference is between direct 

and indirect quotations. 


• content analysis 


Content analysis, as we know, generates quantitative data for further analysis, 

which you might then want to present using traditionally quantitative 

techniques like bar charts, histograms, line graphs and scatter plots. 

Some Final Pointers on the Analysis of Qualitative Data 

These are worth bearing in mind as they will help in any analysis of this kind. 

Do You Need A Transcript? 

A transcript, as we established in Section 6, is a full and complete typed record of all 

of your interview tapes/digital recordings. So to produce a transcript you will need 

to have recorded the interviews themselves, unless you have very good shorthand! 

A good audio typist should be able to transcribe one hour of interview tape/digital 

recording in approximately four hours, and it will take someone who is less skilled 

much longer. If you choose to pay an audio typist to do your transcribing then this 

automatically breaches confidentiality as we have also suggested in Section 6. 

Producing a transcript yourself is also a good first step in qualitative data analysis 

anyway as you have to listen so carefully to what is being said, and it also means 

that your record of the data is complete. Nonetheless you may prefer to listen to the 

recordings and only transcribe certain sections for use in direct quotations whilst 

summarising the rest. Remember here to take down the numbers on the tape/digital 

counter whenever you decide to transcribe a particular section, so you can always 

find the relevant place in the data set. Transcribing machines also make the job of 

transcribing easier as they come with headphones and a foot pedal which allows you 

to stop, pause, rewind and fast forward the recording without taking your fingers off 

the keyboard. 

The Kolb Learning Cycle 

Analysing qualitative data can in many ways be likened to Kolb’s (1984) learning 

cycle. here you begin with a concrete experience (this is the actual data gathering), 

then move to reflective observation (where you familiarise yourself with the data), 

then to abstract conceptualisation (where you identify how the data relate to the 

existing literature) and finally to active experimentation (where you check your 

analysis again to make sure it stands up and possibly, as suggested above, ask your 

respondents to comment on it as well). Active experimentation may suggest that you 

need to gather more data. hence you would return to the concrete experience stage 

of the cycle. Figure 8.1 below illustrates the process in diagrammatic form. 



Figure 8.1: Kolb’s learning cycle as applied to qualitative data analysis 

How Do You Know? And So What? 

It is crucial that any data analysis, whether of primary or secondary data, quantitative 

or qualitative data, answers both of these questions. The first question requires 

that, when you make a claim about your data (e.g., ‘Many respondents thought X’), 

you substantiate it by direct reference to the relevant data themselves. The second 

question requires that you don’t just report or describe your data in this way, but that 

you also analyse them. In other words, this means that you must make the relevant 

links to the literature – do your data support, amend or contradict what is claimed 

there? – as well as suggesting what these data have to say about your research 

questions. 


� Tasks 










Summary 


i) human beings have specific limitations as data analysts – you should try and 

avoid the resulting errors as far as possible. 

ii) Your strategy for analysing qualitative data should depend on the purpose of 

your research. Is it descriptive, explanatory or exploratory? 

iii) Both deductive and inductive strategies have strengths and weaknesses, and 

there are variants of each. Content analysis can be used for both strategies. 

iv) Software is available for analysing qualitative data but should only be used 

when you are sure that you could not do the job just as well manually and/or 

with a basic word-processing or spreadsheet package. 

v) There are various ways to present qualitative data, the most conventional of 

which is the summary narrative plus direct quotations form. 

vi) Finally, consider issues around transcripts, Kolb’s learning cycle and whether 

you can successfully demonstrate your claims about your data and then link 

them to your research questions and the relevant literature. 

References 


University Press chapters 12, 22 and 23 


London: Sage chapter 8 (also see the first edition, 1991, for the full version of the 

content analysis coding sheet example give above) 

Glaser, B.G. (1978) Theoretical Sensitivity: Advances in the Methodology of Grounded 

Theory Mill Valley, California: Sociology Press 

Glaser, B.G. (1992) Basics of Grounded Theory Analysis Mill Valley, California: Sociology 

Press 

Glaser, B.G. and A.L. Strauss (1969) The Discovery of Grounded Theory: Strategies for 

Qualitative Research London: Weidenfeld and Nicolson 

halme, M. (2002) ‘Corporate environmental paradigms in shift: learning during the 

course of action at UPM-Kymmene’ Journal of Management Studies 39(8):1087–1109 

Kolb, D.A. (1984) Experiential Learning: Experience As The Source of Learning and 

Development Englewood Cliffs, CA: Prentice hall 




Practitioner-Researchers 2 nd Edition. Oxford: Blackwell pp. 386–389, chapter 14 



Smart, S. (1994) The Nature of Strategic Change Towards Human Resource 

Management unpublished MBA dissertation, University of Portsmouth 

Strauss, A.L. and Corbin, J. (1998) Basics of Qualitative Research 2 nd Edition. Thousand 

Oaks, California: Sage 






section 9 

Writing Up Your Professional 

Management Project




SECTION 9 

Writing Up Your Professional Management 

Project 



• understand how important it is that your project is well written, 

and how you can achieve high standards in producing written 

text 

• identify the ‘dos’ and ‘don’t’s’ of good project writing practice 

• know where to go for additional help with your project. 

Introduction 

It is worth pointing out from the outset that most people who have grown up 

speaking English are more fluent when they speak than when they write. So the 

advice in this section applies to all of you, not just to those who speak and write 

English as a second (/third/fourth etc.) language. We will cover the importance of 

good writing, offer some advice on how to achieve it and point you in the direction 

of resources which provide additional assistance. 

The Importance of Good Writing In Your Project, and How to 

Achieve It 

Your project mark is entirely based on your written performance, so even those of you 

doing more quantitative research still need to be able to express yourselves clearly in 

writing. What you know is invisible to your assessors unless you can communicate 

it satisfactorily in writing. There is no oral examination/viva for Professional 

Management Projects at the University of Leicester School of Management. 

Writing well involves turning your ideas into a logical and credible argument – and as 

a form of communication it is very different from talking. In other words, when we 

talk to someone the other person can hear us but also has access to our expressions, 

gestures, tone, emphases etc. They can usually also ask for clarification. Readers 

cannot. So you need to ensure that your project is comprehensive, coherent and 



persuasive. It should include all the relevant material as discussed in Section 1 (making 

it comprehensive). It should also be clearly structured, well argued and evidenced and 

focused on its subject matter (making it coherent). Finally, it needs to be sufficiently 

detailed to allow your assessors to see how you arrived at your conclusions and 

recommendations, and therefore to judge them (making it persuasive). Also refer 

back to Section 1 on what makes a good project, some of which will be repeated 

here. 

Learning to write well involves, first, reading others’ work. So think about academic 

authors whose work you find easy to read and understand, because it is accessible 

and logically structured. Then consider the techniques and approaches they use. 

Writing well is all about communicating clearly so you should try and learn from 

others’ successes (and indeed failures). Second you need to observe academic codes 

and conventions, like structuring every chapter using an introduction, discussion 

and summary format (earlier we called this ‘topping and tailing’), referencing all 

your sources properly, explaining technical concepts and so on. Third, you need to 

practise. The more you write – and the more notice you take of our suggestions and 

comments on your previous PDM assignments – the more you will improve. 

Another useful tip for putting together a really good final version of your project is 

to write up as you go. After all, “Writing is first and foremost analysing, revising 

and polishing the text. The idea that one can produce a ready-made text right away 

is just about as senseless as the cyclist who never had to restore his or her balance” 

(Alasuutari 1995:178). In other words, don’t start writing up shortly before you are 

due to submit the project proper. We suggested in Section 6 that by the time you get 

to what is usually called the ‘writing up’ stage, you should already have a full draft 

version of the project which you can then revisit, revise and proofread as necessary. 

Drafting parts of the project as you go makes your life a lot easier when it comes to 

producing the final version. 

Something else you will undoubtedly experience during the writing of your project, 

perhaps at several points, is writer’s block. This is the feeling that you just can’t 

think of anything to say, the horror of the empty screen or page in front of you, a 

lack of motivation to start a new chapter or a redraft of an existing chapter. So here 

are some tips when you just can’t seem to get (re-)started. The first three ideas are to 

draft your contents page, type out your bibliography or type out the quotations 

(from the literature and/or from your data) which you plan to use in the project. 

All three of these tactics mean you have at least put pen to paper or fingers to 

keyboard as well as being useful tasks in and of themselves. Another idea is to draft 

a structure for the chapter you are trying to write. Then you can fill in the gaps 

under each section of the chapter. Another, which many famous authors of fiction 

use, is to set a target for words per day, write up to it then edit. In other words, just 

write until you hit your target and try not to edit as you go. Leave the editing process 

until after you have written the requisite number of words. Otherwise you may find 

yourself endlessly editing the first sentence until you think you’ve got it right, and 

never getting any further! 



You could also try speaking your ideas out loud, recording and transcribing them. 

This allows you to get something down on paper to work with and edit. A variation 

on this theme is to present your ideas to somebody else. Again you can record this 

presentation as well as getting feedback from the other person as to what makes 

sense, what doesn’t, what is omitted, what probably doesn’t need to be there and 

so on. Or try writing at a different time of day. Some of us find it easiest to write 

in the late afternoon and early evening. Other people are better at writing in the 

morning. Experiment to see when you tend to be most productive. 

Before we move on to some technical pointers for better writing, we want to say two 

more things. First, keep your research questions in mind throughout the writing 

process. As we have said with regard to taking notes for and writing up a literature 

review in Section 2, when writing any part of your project try to remember your 

research questions and therefore to ensure that everything that goes into the finished 

product has a direct bearing on those questions. In particular avoid packing your 

project with material that is only vaguely relevant. The temptation is always to put 

in everything that you have read in the subject-specific or methodology literature, or 

experienced whilst planning and carrying out your methodology, but this can lead 

to irrelevancies. Writing well is as much about what to leave out as what to include, 

especially as you only have 6000 words. 

Second, you also need to remember that the project is not just a literature review or 

a description of what is going on in a particular part of the world of management. 

So avoid simply writing up everything you know about organisational structure, 

organisational culture, relationship marketing, charismatic leadership, stress, 

productivity, employee motivation, the supply chain, auditing or whatever else, and/ 

or just describing what your findings were. Remember, as established at several points 

in this module book, academic research is about asking and answering questions 

via a combination of literature and primary/secondary data or a synthesis and/or 

evaluation of (different bodies of) literature. 

Now to some much more specific advice on writing. 

Some Technical Pointers, or the ‘Dos’ and ‘Don’ts’ 

of Writing a Project 

This advice is given so that, if you follow it, your project will communicate as clearly 

as possible to your assessors. First, what should you make sure that you do when 

writing your project? 

The ‘Dos’ of Writing a Project 

Do make sure you provide a clear route map, and that you ‘top and tail’. As suggested 

in Section 1, a good project should provide a chapter-by-chapter overview or route 

map in the introduction so the assessors know what to expect. Each chapter should 

also begin with a short introduction telling the reader what the chapter will do, 



and then end with a short conclusion summarising what has been done, as well as 

signalling what comes next (i.e., reminding the assessors of the route map). 

Do make sure you structure your project logically. You need to ensure that each and 

every point you make follows from the previous point so your argument is clearly 

structured. When you need to shift your discussion into a different topic area, try 

using constructions like ‘Moving on’ or ‘On a different issue’ at the start of the new 

point. This should also be the start of a new paragraph – see below. Also make 

sure you link the sections of the argument in each chapter together, so that your 

project is not just a series of unrelated observations. Linking points make your 

writing flow better and signal the way your argument is developing – e.g., ‘having 

discussed the phenomenon of X, I will now move to analyse the possible explanations 

for this phenomenon’. Subheadings help to clarify an unfolding structure and 

developments in the argument to an assessor as well – but don’t over-use them as 

very short subsections are irritating and distracting. Overall, then, make sure your 

work is well structured in terms of the internal logic of each paragraph as well as the 

links between paragraphs and subsections. 

Do pay attention to paragraphing. Students often tend to make their paragraphs too 

short, which gives projects a ‘bitty’, ‘fragmented’ feel. It is difficult to be prescriptive 

about paragraphing but the usual guidance is to end a paragraph when you have 

finished discussing a particular point. If you feel that your paragraph is getting too 

long (say more than ¾ of a page as a rough guide 32 , and you haven’t come to the 

end of your point yet, then break off at a logical place and restart using a construction 

like ‘Moreover’, ‘Furthermore’, ‘Additionally’ etc. These sorts of constructions make it 

clear that you are still discussing the same issue. 

Do be careful about forms of language. Certain forms of language may send unwitting 

or unintentional messages – e.g., the use of gender-specific/exclusive language like 

‘he’, ‘him’, ‘his’ etc. This may be grammatically correct but it is often not empirically 

or politically correct! We prefer you to use gender-inclusive language like ‘he/she’ 

or gender-neutral language (e.g., ‘they’) unless you are referring to someone whose 

gender you know or quoting someone who uses gender-specific language. Insensitive 

or outdated language should also be avoided – e.g., ‘person of colour’ is preferable 

to ‘coloured person’, as suggested in Section 5. And consider the use of terms like 

‘research subject’ when talking about those who took part in your research. Might 

this not signal that you ‘subjected’ them to something or that they are subordinate or 

inferior to you in some way? This connects back to our discussion of research ethics in 

Section 6. We would recommend the use of less problematic terms like ‘respondent’, 

‘informant’, ‘participant’ and so on. 

Do use either the first person or the third person, as illustrated in figure 9.1. 

32 Bear in mind that the text in the project needs to be at least 1.5 spaced as 

well, so this advice about paragraph length refers to the finished version of the text. 

The Project Guidelines give full details of required formatting. 



First person version Third person version 

My project focuses on … This project focuses on … 

I would argue that … It can be argued that … 

My interpretation of this concept is … One interpretation of this concept is … 

I conducted interviews with … The researcher conducted interviews 

with … 

My respondents suggested that … The respondents suggested that ... 

Figure 9.1: First person usage versus third person usage 

In the past, academic research was usually written up using the third person, as it 

is assumed to sound more objective – as if the piece is more than just someone’s 

opinion. But more recently the first person has become acceptable as well. So we 

would suggest choosing the approach with which you are most comfortable – i.e., 

the usage you have chosen in your assignment writing or the writing you do for work. 

There are two additional things to remember here though. First, writing in the first 

person can sound too ‘chatty’, informal and colloquial if this approach is not used 

carefully. Your project is a formal piece of academic work and needs to be presented 

as such. Second, we would recommend that you use either the first person or the 

third person throughout your project as switching between them can be distracting 

for the assessors. 

Do be careful about the use of footnotes and endnotes. Notes are used to expand 

on an argument, to include material which might make the text seem unbalanced if 

it was included in the main body of the argument. Most word-processing packages 

allow you to insert footnotes (which appear at the base of the relevant page, as 

in this module book) and endnotes (which appear at the end of the document) 

automatically. however, as with appendices (see Section 1), try to keep notes to a 

minimum. If the material is really so important it probably belongs in the main body 

of the text. Also, always use either footnotes or endnotes throughout, not both. 

Do pay attention to the use of the past tense and the present tense. One academic 

convention is to use the present tense throughout (e.g. ‘Weber claims that’ not 

‘Weber claimed that’) except when writing up your methodology and data analysis 

where you should use the past tense (e.g. ‘I chose to use questionnaires’, ‘my data 

suggested that’). But these are not hard and fast rules, and it may be a matter of (a) 

what works for you; and (b) what makes the project most ‘readable’. In any case, 

please try to avoid swapping between tenses within chapters. 

Do cite all your sources. Remember to cite the sources of all the indirect and direct 

quotations you have used in the main body of the text. You also need to provide a 

full list of references/bibliography at the end of the project, before any appendices 



– and don’t try and pad this out! Only list sources you have actually consulted, 

and remember that if the list seems excessively long then your assessors might get 

suspicious! Students also sometimes ask whether they should include sources in the 

list of references which they haven’t cited in the main text. If you really think that 

the source in question has informed your thinking then include it in the list. If not, 

leave it out. And, again, remember that lots of sources that don’t appear in the text 

but do appear in the list of references will be something which generates negative 

commentary from your assessors, because again it looks like padding. 

Do be careful about the use of direct quotations. Direct quotations prove you have 

read the source in question and allow assessors to see how competently you have 

used it. But don’t over-use direct quotations and certainly do not use lots and lots 

of long ones (anything above 40 words or so). Further, don’t make the mistake of 

piling direct quotations on top of each other, so that one follows from another and 

then another with none of your own words in between. In fact the general rule is 

that, if you can say it just as well yourself, then don’t use a direct quotation. Use an 

indirect one instead. In general we are interested in your interpretations of other 

people’s ideas and lots of direct quotations don’t really allow us to see how well you 

understand the material you have used. IN ANY CASE, MAKE SURE ThAT YOU PROVIDE 

AN APPROPRIATE REFERENCE WhEN YOU ARE QUOTING DIRECTLY FROM A SOURCE 

TEXT (AS WELL AS USING QUOTATION MARKS TO DENOTE ThE RELEVANT PASSAGE) 

AND REMEMBER ThAT INDIRECT QUOTATIONS ALSO NEED TO BE REFERENCED. MORE 

ON ThIS ISSUE BELOW IN ThE DISCUSSION OF PLAGIARISM. You might also want 

to indent longer direct quotations and/or present them using single spacing to 

enhance readability. 

Do cross-reference where necessary. In other words, you should do what we have 

been doing throughout this book – signal to the reader where something will be 

discussed in more detail or from another angle later in the project, and refer back to 

previous discussion to remind them of an argument where necessary. 

And now to the things you should avoid doing in your writing! 

The ‘Don’ts’ of Writing a Project 

Most of these should be obvious from what we have already said. As with the ‘dos’ 

above, they also apply to other academic assignments. 

Don’t provide an unclear or inadequate outline of your research questions. Make 

sure that your project clearly spells out your research questions and why you are 

asking them in the introduction, and makes appropriate reference to these questions 

and justifications in later chapters where necessary. If the assessors aren’t clear on 

what your focus is from the outset, and why this focus has been chosen, this will have 

a negative impact on your mark. 

Don’t commit the errors of poor referencing or plagiarism. Plagiarism can be from 

published texts of whatever sort (including Internet sources) or from other students’ 

work. AS YOU KNOW BY NOW, EIThER WAY IT WILL ATTRACT A FAILING MARK. MAKE 



SURE YOU REFERENCE PROPERLY AND REMEMBER ThAT, IF YOU DON’T UNDERSTAND 

A PARTICULAR TEXT, ThEN YOU NEED TO PUT MORE EFFORT INTO READING IT AND/ 

OR TO DISCUSS IT WITh A TUTOR ON BLACKBOARD. DON’T JUST COPY ChUNKS OF IT 

OUT. WE WILL IDENTIFY PLAGIARISM AND IT WILL LEAD TO A FAILING MARK – PLEASE 

BE WARNED! Equally, inconsistent or incomplete referencing will also affect your 

mark – and not in a good way! 

Don’t ‘jump around’. This means writing in a series of disconnected points so the 

reader cannot see how the argument or the storyline of the project builds. You 

should be telling a coherent narrative. 

Don’t make unsupported claims or offer too much personal experience as evidence. 

If you make a claim about something – e.g. ‘Money is the best motivational tool 

managers have available’ – then you need to be able to provide evidence of this claim. 

Evidence can be from theory, from empirical data (your own or someone else’s) or 

your personal experience. however, don’t overuse your own experience as this makes 

a project sound too anecdotal – i.e., like a series of personal stories as opposed to a 

properly researched piece of academic work. Concrete, ‘real life’ examples of what 

you mean are also good – e.g., a description of how a financial incentive scheme at 

Organisation X raised productivity by 100% in less than six months, as discussed by 

Author Y (2006) to illustrate the above claim about money as a motivator. In general, 

when you make a claim, always ask yourself ‘how do I know this is the case?’ and 

provide the relevant evidence. Also make sure that the evidence you provide fits the 

claim you are making! 

Don’t overstate the case, neglect the other side of the argument or ignore 

critiques. Overstating the case or neglecting the other side of the argument are both 

problematic as most if not all topics in the management literature are characterised 

by controversy and disagreement. We touched on this issue in Section 2. So try not 

to exaggerate the strength of one particular way of understanding an issue – avoid 

presenting it as definitive. here is an example of what we mean: 

’Effective organisational learning is the most important source of 

corporate success’ 

versus 

‘Effective organisational learning is widely regarded as an important 

source of corporate success’ 

The first claim exaggerates the importance of organisational learning – the second is 

more accurate and nuanced. 

In terms of ignoring critiques, and again because of the discussion and debate that 

exist in nearly all areas of the academic literature on management, try not to present 

one author or one perspective as if they are in some way the ‘Gold Standard’. An 

example here is the way in which a lot of management students we have come 

across use the work of Geert hofstede on cross-cultural differences in organisational 



behaviour. As we suggested in Section 2 hofstede is one of the key authors in this 

area of the literature. however, the mistake students often make is to present his 

work as if it is the be-all-and-end-all, as if there is nothing else to say on the subject. 

There is no problem at all with a project that makes a lot of use of hofstede – but it 

would need to acknowledge that his ideas are very controversial and widely criticised, 

as well as to make a robust case as to why the student in question has chosen to use 

his work. 

Don’t say too little. This often happens when students try to fit too much into their 

word limit, and don’t explain key points fully enough. Assessors then have to read 

between the lines and guess at what you mean, which will affect your mark. Don’t 

leave arguments ‘hanging in mid-air’. Make sure all your points are appropriately 

developed so that they are clear. As we have said above, writing well means knowing 

what to leave out as well as what to include, and the key balance is between breadth 

and depth of coverage. 

Don’t include irrelevant material. Relatedly, everything that goes in should have 

some bearing on your research questions (as we have also said above). It should 

be part of the unfolding story which we talked about in Section 1. Make sure your 

assessors are never left wondering why you are talking about a specific issue. 

Don’t write in an informal style or use inappropriate language. Avoid using slang 

or colloquialisms. Avoid overusing contractions like ‘don’t’ and ‘didn’t’. Avoid 

writing in a chatty or informal way as if you were writing an e-mail or text to a friend. 

Finally, avoid using sexist, racist or homophobic language (etc.) – unless of course you 

are quoting another author or a respondent who uses these kinds of constructions. 

Don’t use jargon without defining it. If you are using technical terms or jargon which 

belong to a specific area of the academic literature, and would not be understood by 

someone who is not familiar with this literature, then you need to define these terms. 

Otherwise we don’t know whether you understand them or not. The trick here is to 

write for an intelligent layperson – i.e., someone who is not an expert in the area 

but will understand the relevant issues if they are carefully explained. As assessors we 

will of course be familiar with the material you are discussing, but what we want to 

see is that you understand it. 

Don’t overwrite. Effective communication is about being clear, concise and crisp. 

Although some academics use very complex and lengthy sentences and extremely 

flowery and elaborate language, this does not mean that academic work needs to 

have these characteristics. An example of overwriting follows: 

“While it is true that researchers have illusions of academic grandeur 

when they sit down to write their project report, and who can blame 

them because they have had to demonstrate skills and resilience to get 

to this point in their studies, they nonetheless must consider that writing 

a project is an exercise in communication and nobody likes reading lots 

of ideas which are expressed in such a confusing and pretentious way 



that nobody can understand them, let alone the poor tutor who has to 

plough through it all to try and make some sense of it.” 

Phew!! This is an instance of using sentences which are much too long. What follows 

is much more carefully expressed: 

“Researchers have illusions of academic grandeur when they write their 

project report. This is understandable. They have demonstrated skill 

and resilience to reach this point in their studies. however, writing a 

project is an exercise in communication. Nobody likes confusing and 

pretentious writing which is difficult to understand. We should pity any 

tutor who has to make sense of it.” 

This is a much clearer rendition of the same points – and cuts up one sentence into 

six. A tip to avoid very lengthy sentences is to read your work out loud. The points at 

which you need to breathe are usually where a full stop/period is needed. 

Don’t skimp on your research. Not reading enough, whether it’s subject-specific or 

methodological literature, makes your argument weak, as does overrelying on some 

sources – plus it looks lazy! 

Don’t write a conclusion that contradicts your main discussion. Make sure your 

conclusion doesn’t contradict the discussion in the preceding chapters. We have seen 

lots of examples where project conclusions reflect what the student wanted to find 

as opposed to what their research actually found. Inappropriate generalisations 

from a non-representative sample should also be avoided, as should sweeping and 

unrealistic recommendations which are not substantiated by your data (also refer 

back to Sections 1, 7 and 8). 

Don’t submit a project which contains typographical errors. An untidy project with 

lots of errors will not achieve a good mark. Make sure your plan leaves time for you 

to proofread the project yourself (again reading out loud helps) or to get someone 

else to do it (as long as this is all they are doing – they shouldn’t be advising you on 

content). Also 

1. check to see that you have followed all the formatting instructions in the 

Project Guidelines; 

2. try to ensure your title is (a) relevant to the project, (b) informative in terms 

of letting assessors know what to expect and (c) exciting/attention grabbing; 

3. check your contents list to make sure it’s comprehensive and accurate; 

4. check to make sure all the pages are in numerical order; 

5. do a spellcheck and a grammar check; 

6. and, finally, check your referencing. 



Don’t write too many words or too few. here are some ballpark figures for each 

chapter: 

introduction: 750 words 

literature review: 1500 words 

methodology: 1500 words 

data analysis: 1500 words 

conclusions, recommendations and reflections: 750 words 

We mean what we say when we use the term ‘ballpark’. Remember that a project 

should be 6000 words long, but that you are permitted 10% leeway above and below 

this figure (i.e., from 5400 words to 6600 words as a minimum and maximum). So 

the figures above should be seen from this point of view. Some projects will also look 

rather different from this set of figures for word count per chapter depending on 

their subject matter. however, very long or very short projects will always attract poor 

marks. Anything preceding the introduction (e.g., your executive summary/abstract) 

and anything following the reflections section (e.g., appendices and bibliography) 

are not part of the word count. But remember that this does not mean you should 

dump lots of material in the appendices because you can’t fit it in the main text! 

To sum up, what follows is a very rough guide to what we regard as some of the 

major, medium, minor and trivial ‘sins’ in a project. We will take account of any such 

sins in our marking, but obviously lots of major sins are likely to have a much more 

negative impact on your mark than a few minor sins: 

Major sins: reproduction of others’ words and ideas without acknowledgement (i.e., 

plagiarism); project is not clearly focused; project is incoherent, badly written and/ 

or poorly structured; project demonstrates inadequate engagement with relevant 

literature, both subject-specific and methodological; project is one-sided; project is 

derivative or uncritical; project contains extremely poor or no referencing; project 

omits or neglects key components (e.g., methodology). 

Medium sins: significant omissions from the references, or referencing is inconsistent; 

project is not spellchecked – there are a considerable number of incorrectly spelt 

words; project is poorly structured. 

Minor sins: project is poorly punctuated; project contains some misspellings; project 

omits some references. 

Trivial sins: project contains a few misspellings; project demonstrates some poor 

punctuation; project uses et al. in the list of references when all authors should be 

given at this stage. 



And, finally, where else can you go for support with your project, including writing 

it up? 

Other Sources of Advice and Information on 

Writing Your Project 

here are some more sources of help with regard to writing up your project. The 

first is the ULSM assignment writing guidelines which you should be familiar 

with by now. You can find these at: http://www.le.ac.uk/ulmc/existingstudents/ 

assignwritingguidelines.pdf. Check these out for guidance on preparing, writing and 

referencing assignments (including the project) correctly. Some of it is material we 

have adapted for this section, but some of it will be new to you. Second, at http:// 

www2.le.ac.uk/offices/ssds/sd you will find the home page of Student Development 

at the University of Leicester. There are all sorts of useful resources here on writing, 

studying and researching. 

The Project Guidelines – of course – should be read, digested and committed to 

memory. To remind you, the Guidelines can be found on the Professional Management 

Project site on Blackboard. Also on Blackboard, as noted earlier, you will find the 

project Discussion Boards. Just as with the other modules you have studied during 

your Diploma programme, there are discussion boards where you can ask questions 

of a ULSM tutor relating to issues about your project. This is how you receive your 

project supervision. The boards are as divided by topic area follows: 

Accounting and Finance 

Marketing 

Management (incorporating Strategy, Organisational Behaviour and 

hRM) 

Now, some tips about how to play nicely on the discussion boards: 

1. Read the Project Guidelines and this module book before posting – so you 

know basically what we expect and what you need to do. Also check other 

posts on the board to see if your question has already been answered. 

2. Be clear and succinct – keep your posts short and sweet (no more than 500 

words) and don’t ask the tutor to read draft chapters or sections of chapters, 

as we are not able to offer this service. 

3. Choose the right board – in other words if you are doing a Marketing project 

then choose the Marketing board. Otherwise the tutor may redirect you to 

another board. Please also do not post to more than one board. 



4. Do not begin work on your project proper until you have discussed it with a 

tutor via one of the boards. Equally, do not begin work on your data collection 

until you have received tutor approval for your proposal via the board, as 

discussed in Section 6. 

5. Remember this is not a confirmation process – tutors will comment on your 

ideas, offer support and advice and approve your proposal but you should 

not assume any of this means your project will receive a passing mark. It 

is up to you to take the feedback you are given and turn it into a decent 

submission. 

6. If posting to a discussion board for the first time the minimum information we 

expect from you is the research questions you want to ask, the subject-specific 

literature you plan to use and an outline of your proposed methodology if 

relevant. If you don’t supply this information it often means that the tutor 

can’t answer your question. 

7. Observe the three working days rule – we guarantee a response on the 

board to your query within three UK working days (i.e., Monday to Friday), 

unless the tutor has notified you via the board that they are away for a period 

of time (see point 8 below). Please do not ‘hassle’ the tutor for a response 

unless this time period has elapsed. 

8. Don’t post when the tutor has asked you not to – tutors need holidays too. 

At times they will ask students not to post to a board while they take a wellearned 

break, so they don’t return to piles and piles of posts. This is especially 

likely during the UK summer or at Christmas. however where possible we will 

arrange cover for a board during tutor holidays. 

9. Don’t try to take discussions off the board or spam tutors – we will not 

respond to spam e-mails sent using the BB communication function, and will 

simply redirect you to the relevant board. Also unless there is a very good 

reason we will only provide supervisory assistance on BB, not by individual 

e-mail. This is because students often ask the same question and it makes 

much more sense for everyone to be able to see a discussion thread on that 

basis. 

Finally, remember that all ULSM tutors mark many student research projects every 

year. So the clearer, more enjoyable and more interesting your project is, the better 

your mark will be. A tip is to ask a friend to read your executive summary/abstract or 

your introduction to see if it inspires them to read further … 

Best of luck with your project! 




� Tasks 

Summary 









1. Writing well means being comprehensive, coherent and persuasive. 

2. You can learn to write well by reading others’ work, following academic codes 

and practising. 

3. Write up your project as you go. 

4. Keeping your research questions in mind is extremely important. 

5. Also pay attention to technicalities like paragraphing and forms of language. 

6. There is a long list of ‘don’t’s’ in projects such as not defining the jargon you 

use and overwriting. 

7. Don’t forget to consult the ULSM Assignment Writing Guidelines, the Student 

Learning Centre resources and the Project Guidelines, as well as posting any 

queries about your project to the relevant BB discussion board. 

8. Remember to keep your assessors entertained! 

References 




Alasuutari, P. (1995) Researching Culture: Qualitative Method and Cultural Studies 

London: Sage 







Saunders, M., P. Lewis and A. Thornhill (2009) Research Methods for Business 

Students 5 th Edition. harlow: Financial Times Prentice hall chapter 14, appendix 4 

(pp. 584–586) 




aPPenDix 1 

Doing a Theoretical Project




APPENDIX 1 

Doing a Theoretical Project 

As we have stated already, in our experience most students of management opt to 

do an empirical project, dissertation or thesis to achieve their qualifications. It is 

important to remember that using secondary data – i.e., data that has been collected 

by someone else but is available to you, such as market reports, government census 

data or organisational data like financial accounts or absence statistics – is still 

empirical research. In other words, it involves using someone else’s data to answer 

the research questions you have posed. As such projects which only use secondary 

data should be structured like a standard empirical project as outlined in Section 1 of 

this module book. 

however, some of you may prefer to undertake a theoretical project, sometimes 

referred to as a library or desk-based project. We have provided the following guidance 

for those who decide to go down this route. Please remember the following before 

you embark on this kind of project: 

1. Doing a theoretical project is not an easy option. Indeed it is often more 

straightforward to choose an empirical project. This is because working with 

empirical data usually makes it easier to say something new about a subject 

area as opposed to simply rehashing what the relevant academic literature 

suggests about that subject area. Remember that research is about answering 

questions, seeking clarification on things we are not sure about and clearing 

up gaps in the breadth or certainty of our knowledge about the social world. 

The term ‘research’ actually implies uncertainty. It does not involve telling the 

reader something that they already know, or could find out if they read the 

relevant literature for themselves. 

2. Relatedly, doing a theoretical project means you have to summarise the 

existing literature(/s) in a subject area(/s) in a way that adds something to 

that literature. A straightforward literature review where you simply tell us 

about the key arguments and debates is not sufficient. You must evaluate the 

literature as it stands and/or synthesise (bring together) more than one body 

of literature to see what one can add to the other or where there might be 

similar or different sorts of claims being made. 

3. It is equally important to check your initial ideas out on the relevant Blackboard 

project discussion board when pursuing a theoretical project as it would be if 

you had chosen to do an empirical research project. And you are still required 

to submit a one page research proposal in this case as well. 

4. If you undertake a theoretical project, then the outcome should be a series 

of recommendations for further academic research as opposed to practical 

recommendations for organisational change or managerial initiatives as in 

many empirical projects. The idea is to look for gaps, weaknesses, problems 



or biases in the existing literature – to undertake a critical analysis – in order 

to lay the ground for future research. 

5. Theoretical projects demand that you provide a very thorough review of 

the existing literature/s. In a standard Diploma empirical project, the three 

substantive chapters – literature review, methodology and data analysis – 

should come in at somewhere around 1500 words each. For a theoretical 

project, these three chapters become an extended literature review, divided 

appropriately, and running to a total of some 4500 words. This means that 

you will need to read more of the existing literature than you would if you 

were doing empirical research. 

Some possible routes for theoretical projects might include one or a combination of 

the routes below: 

• An analysis which suggests that the existing literature is dominated by 

commentary which focuses on one particular part of the world. It is common, 

for example, for management literature to be written by academics from the 

Global North, about issues in the Global North, using data from the Global 

North. Thus organisational and management issues which are specific to 

the Global South are relatively under-discussed. This might also mean that 

management theories and concepts which are not applicable to the Global 

South are being taught and disseminated there even though they do not ‘fit’ 

properly. 

• An analysis which suggests that the existing literature on a specific topic 

is dominated by a particular methodology – for example, the use of selfadministered 

questionnaires on representative samples – which might offer 

a one-sided view of the topic being investigated. This might then lead into 

an analysis which proposes that a more qualitative, open-ended, nonrepresentative 

strategy could bring different aspects of the topic to light. 

• An analysis which establishes other conceptual or empirical gaps in a body of 

literature. For example, there is a large body of literature on work-life balance, 

but the vast majority of this literature concentrates on how the struggle to 

manage this intersection is something that mainly affects women and/or 

parents. There is very little which focuses on men and/or those who have 

other sorts of dependents – e.g., elderly parents – and/or other life interests 

and commitments. This has the effect of narrowing discussions of work-life 

balance initiatives and recommendations, as if only a specific proportion of 

the workforce worldwide are trying to juggle ‘work’ and ‘life’. 

• An analysis which brings together or synthesises more than one body of 

literature with the intention of suggesting what one set of theories or claims 

might add to or have to say about the other, or to establish that there might 

be more common ground between two competing camps than has at first 

been thought. A similar strategy is to compare an emerging literature against 

a body of thought which it claims to replace or improve upon. An example of 



this strategy is the commentary on human Resource Management during the 

1990s which suggested that it was either something genuinely new or in fact 

that it was not very different from old-style personnel management and was 

therefore a case of ‘old wine in new bottles’. 

here are some examples of published theoretical research papers by School of 

Management faculty. They are provided to illustrate what we mean by saying 

something new about, evaluating or synthesising existing literature/s. We suggest 

you read some or all of these papers yourself to get a better sense of what theoretical 

research entails. 

Brewis, J. (2001) ‘Foucault, politics and organizations: (re)-constructing sexual 

harassment’, Gender, Work and Organization 8(1):37–60 

here Jo Brewis evaluates the academic literature on workplace sexual harassment. 

She uses the ideas of the French philosopher Michel Foucault, who argues that 

various forms of knowledge – like academic literature – combine to provide ideas, 

norms and frameworks which we as human beings use as a guide in how we live 

our lives. In other words, Foucault suggests that knowledge has ‘power effects’ – it 

is powerful in constructing the ways we understand ourselves and each other. Jo 

argues, on this basis, that the existing literature on/knowledge of sexual harassment 

might have several unintended consequences which actually serve to reproduce 

harassment in the workplace, instead of helping to eliminate it. For example, she 

suggests that the argument in the literature that women are the helpless victims of 

sexual harassment, whereas men are the aggressive perpetrators, serves to reproduce 

the prevailing assumption that men should initiate sexual contact – almost to the 

extent of suggesting that women are so sexually passive that they can’t even fight 

back when they are being harassed. here then Jo evaluates the literature on sexual 

harassment by using Foucault’s arguments. So she is also synthesising two literatures 

– bringing Foucault’s ideas to bear on the harassment literature. 

Dunne, S. (2008) ‘Corporate social responsibility and the value of corporate 

moral pragmatism’ Culture and Organization 14(2):135–149 

Stephen Dunne starts by suggesting that academic discussions of Corporate Social 

Responsibility (CSR) increasingly focus on what CSR should be – what it should look 

like in practice – as opposed to discussing whether it should be/exist – i.e., is it a 

good idea at all? his paper then goes on to ask whether this pragmatic focus on 

how we should go about doing CSR is something that should be welcomed. Stephen 

evaluates the central CSR claim that corporations can be regarded as moral actors 

or as having a moral ‘personhood’, rather like individual people. he contrasts the 

work of Milton Friedman – who opposes this claim – to the work of those who 

accept it. Stephen suggests that those in the latter camp, who support the idea 

of CSR, tend to discuss how to operationalise it (i.e., to define and/or put it into 

practice), rather than engaging properly with Friedman’s critique. however, Stephen 

also states this is not necessarily because he agrees with Friedman. Instead he argues, 

using Nietzsche and hanlon, that a proper critique of Friedman must also critique his 

fervent belief in free market capitalism – which the advocates of CSR seem reluctant 



to do. Thus these advocates are closer to Friedman than it may at first appear. here 

then Stephen suggests that these two apparently opposing camps actually share 

some common ground, and he also highlights areas that still need to be addressed 

in CSR commentary. Again then he is both evaluating the CSR literature and also to 

some extent synthesising its different camps. 

Parker, M. (1999) ‘Capitalism, subjectivity and ethics: debating labour process 

analysis’ Organization Studies 20(1):25–45 

In this paper Martin Parker tries to establish whether there might be any similarities 

between two competing versions of organisational analysis. The first is labour process 

theory, which is broadly Marxist. Labour process theory believes in an essential 

human nature and the possibility of universal human freedom. It is rooted in a 

critique of work arrangements under capitalism as alienating and exploitative. The 

second is poststructuralist analysis which does not believe in essential human nature 

or universal human freedom. Poststructuralism asks us instead to reflect upon the 

way that we currently organise the world around us (e.g., capitalism) in order to seek 

out the alternatives – which may or may not be preferable to the ones currently in 

place. Martin identifies the key differences and similarities between theorists in these 

camps and concludes by suggesting that the common ground they might share has 

to do with ethics. In other words, both camps challenge what ‘is’ and debate what 

‘ought to be’ instead in the organisational context. Again then Martin is evaluating 

claims from each body of literature but also synthesising them. Like Stephen Dunne’s 

analysis, this is an interesting strategy because there is at first glance much more 

disagreement between these two camps than agreement. 


© University of Leicester 2009 

Leicester LE1 7RH 

UK 

www.le.ac.uk

PMP module book final.pdf - Blackboard - University of Leicester

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?