12.01.2015 Views

RESEARCH METHOD COHEN ok

RESEARCH METHOD COHEN ok

RESEARCH METHOD COHEN ok

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

PRACTICAL CONSIDERATIONS IN QUESTIONNAIRE DESIGN 343<br />

<br />

<br />

<br />

<br />

<br />

<br />

<br />

Balance brevity with politeness (Oppenheim<br />

1992: 122). It might be advantageous to replace<br />

abluntphraselike‘maritalstatus’withagentler<br />

‘Please indicate whether you are married, living<br />

with a partner, or single ....’ or ‘I would be<br />

grateful if would tell me if you are married,<br />

living with a partner, or single’.<br />

Ensure a balance of questions that ask for<br />

facts and opinions (this is especially true if<br />

statistical correlations and cross-tabulations are<br />

required).<br />

Avoid leading questions.<br />

Try to avoid threatening questions.<br />

Do not assume that respondents know the<br />

answers, or have information to answer<br />

the questions, or will always tell the truth<br />

(wittingly or not). Therefore include ‘don’t<br />

know’, ‘not applicable’, ‘unsure’, ‘neither agree<br />

nor disagree’ and ‘not relevant’ categories.<br />

Avoid making the questions too hard.<br />

Balance the number of negative questions<br />

with the number of positive questions (Black<br />

1999: 229).<br />

Consider the readability levels of the<br />

questionnaire and the reading and writing<br />

abilities of the respondents (which may lead<br />

the researcher to conduct the questionnaire as<br />

astructuredinterview).<br />

<br />

<br />

Put sensitive questions later in the questionnaire<br />

in order to avoid creating a mental set<br />

in the mind of respondents, but not so late<br />

in the questionnaire that boredom and lack of<br />

concentration have set it.<br />

Intersperse sensitive questions with nonsensitive<br />

questions.<br />

Be very clear on the layout of the questionnaire<br />

so that it is unambiguous and attractive<br />

(this is particularly the case if a computer<br />

program is going to be used for data<br />

analysis).<br />

Avoid, where possible, splitting an item over<br />

more than one page, as the respondent may<br />

think that the item from the previous page is<br />

finished.<br />

Ensure that the respondent knows how<br />

to enter a reply to each question, e.g.<br />

by underlining, circling, ticking, writing;<br />

<br />

<br />

provide the instructions for introducing,<br />

completing and returning (or collection<br />

of) the questionnaire (provide a stamped<br />

addressed envelope if it is to be a postal<br />

questionnaire).<br />

Pilot the questionnaire, using a group of respondents<br />

who are drawn from the possible<br />

sample but who will not receive the final,<br />

refined version.<br />

With the data analysis in mind, plan so<br />

that the appropriate scales and kinds of data<br />

(e.g. nominal, ordinal, interval and ratio) are<br />

used.<br />

Decide how to avoid falsification of responses<br />

(e.g. introduce a checking mechanism<br />

into the questionnaire responses to<br />

another question on the same topic or<br />

issue).<br />

<br />

<br />

Be satisfied if you receive a 50 per cent<br />

response to the questionnaire; decide what<br />

you will do with missing data and what is the<br />

significance of the missing data (that might<br />

have implications for the strata of a stratified<br />

sample targeted in the questionnaire), and why<br />

the questionnaires have not been completed<br />

and returned. For example, were the questions<br />

too threatening or was the questionnaire too<br />

long (This might have been signalled in the<br />

pilot).<br />

Include a covering explanation, thanking the<br />

potential respondent for anticipated cooperation,<br />

indicating the purposes of the research,<br />

how anonymity and confidentiality will be addressed,<br />

who you are and what position you<br />

hold, and who will be party to the final<br />

report.<br />

If the questionnaire is going to be administered<br />

by someone other than the<br />

researcher, ensure that instructions for administration<br />

are provided and that they are<br />

clear.<br />

Akeyissuethatpermeatesthislengthylistis<br />

for the researcher to pay considerable attention<br />

to respondents; to see the questionnaire through<br />

their eyes, and envisage how they will regard it (e.g.<br />

from hostility to suspicion to apathy to grudging<br />

Chapter 15

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!