EE7205: Research Methods
Dr. P. N. Suganthan
Room No No.: : S2 S2-B2a-21 B2a 21
Tel : 67905404
E-mail: E mail: epnsugan@ntu email@example.com edu sg
(Edited from Prof Martin Adam’s slides)
What to Expect at End of Course
• Explain reasons for carrying out research.
• Produce a simple project proposal.
• Undertake a literature search.
• Plan experiments p w.r.t. scientific
• Present data appropriately pp p y and explain p
• Understand requirements q for ppresenting g
research (papers, talks etc)
• Appreciate pp the requirements q for pprotecting g
and exploiting research.
• Learn basic skills to become a successful
• Improve research methodology
experimental design, professional ethics &
academic integrity, oral & presentation
• Perform literature survey (you choose
topic), analyse papers, summarise + write
survey in form of scientific paper.
Technical Writing and
• Introduction to Papers and Related Issues
• The Structure of a Scientific Paper
• How to get your Paper Accepted – the
Reviewers Point of View
• The Process of Submitting a Paper
• Making a presentation
Types of Scientific Papers
• Journal papers
– Letter, 3 to 4 pages
– Full paper, 3 to several tens of pages
• Full papers
– NNormal l papers, which hi h generate t kknowledge l d or new
understandings of known issues
– Review papers
• Conference papers
• Technical reports – not reviewed
– For example, IEEE Spectrum, may or may not
• Books – least common way of publishing results
A Useful Book
• If you need a book to refer to to, you can try
– Th Thomas N. N Huckin H ki and d LLeslie li AA. Ol Olsen,
“Technical Writing and Professional
Communication for Nonnative Speakers of
English”, 2nd Edition, McGraw-Hill, 1991.
NTU library has about 10 copies.
Conference papers p p
• Conference and workshop papers
– Very common in Electrical Eng. and Comp.
Science areas. areas
– You are expected to present your paper paper, don’t don t
submit if you do not intend to go!
– Printed in proceedings.
– Accessible over internet.
Author’s Author s goals
• To get published – to graduate
• To describe interesting new knowledge in an
• Must be unpublished work (except review
papers) p p )
– It may be an extension of your previous work
• You can cite other people’s work, but never,
ever, claim someone else’s work as yours!!!
The editor and reader reader’s s perspective
• Suppose you are an editor
– Is the paper in my journal domain?
– Is there new finding? g
– Will it have a wider audience? – boost my journal
– IIs it understandable? d t d bl ? - EEnglish li h
• Suppose you are a reader
– IIs it new and d relevant l t to t me? ?
– Helpful to my research?
– Easy to read?
Some questions to ask yourself
bbefore f writing i i
• Who is going to read this?
•Why y did I do the work in the first place? p
• What is the problem I have solved?
– What is new?
• Is my work new and relevant?
• What are my contributions?
• Are there remaining unsolved questions?
The reader should be …
• Intrigued within the first 5 minutes of
• Excited within 15 minutes.
• Satisfied after 45 minutes.
A survey paper
• Who is going to read this?
• What is the problem?
• Which method did people use?
• What are the most relevant contributions?
• AAre there th remaining i i unsolved l d questions? ti ?
Dealing g with page p g limits
• Letters usually have very limited pages
• JJournal l can accept llong papers with i h extra
charge (overlength page charge)
• Conferences usually have a page or word
• Don’t cheat with small fonts, narrow
margins g etc.
• Try to eliminate redundant materials
• Trim the technical section section, if necessary
Where to submit
• FFor School S h l of f EEE students, t d t
– IEEE series, for example, IEEE Signal Proc, Trans. On Robotics
– Applied Physics, for example, APL, JAP.
• Decide where to submit before you write
– Different journals have different formats.
– YYou may also l need d tto cater t ffor a particular ti l audience. di
• Journal paper usually needs longer time
– MMay take t k one or two t years.
– Follow “Instructions to Authors”.
• Conference paper – Remember important dates
– Notification: reject or accept.
– Camera-ready copy copy.
– Actual conference with presentation.
• Articles can be rejected for various reasons.
• For example, example your topic does not have a wide
audience (for Nature, Science journal etc.)
• Some conferences accept only 15% 15%.
• However … you get valuable reviews.
• Think twice before complaining (appeal) (appeal).
• Reviewers are anonymous, don’t try to guess
who they are. are
• Use the criticism constructively – if they
misunderstood you you, others will do as well well.
• Consider sending it elsewhere.
Th The Structure St t of f a Paper P
• Must be informative, clear and meaningful.
• Should gain attention of the reader
• Don’t use symbols in titles.
• Imagine someone searching for your paper paper.
A skeleton paper
• Title, author(s) and affiliation(s)
• Index term/keywords
– Background information
– Problem definition
– Summary of contributions
– Related works
• New Ideas/Experiments
• Results and discussions
• Appendix if any
• Summarizes problem, reasons for the
work and the results.
• Short and concise, 100 – 250 words.
• Avoid references and acronyms.
• Try to sell the paper to the reader – why
sho should ld he/she read it!!!
Indexing and Keywords
• Use them to classify the paper
• A few categories that I use often:
– Differential Evolution
– Ensemble Methods
– Artificial intelligence
– GA (abbreviations ( should be well known!) )
– Index Terms or Keywords (IEEE)
Background g information
• A short introduction to the area.
• What is the current interest?
• What is the current development/status?
• Cit Cite relevant l t references f and d maybe b bbriefly i fl
explain what has been reported there
• Explains the background/significance of
• Th The opening i paragraph h should h ld bbe
• Ended by y what yyou
have done in this
– In this paper, we shall present ……
• A concise statement of what you are
• Why y it is useful to do/solve?
• Identifying a loophole/over looked issue issue.
• Eg: Current liquid crystal displays have a
slow response - results in blurred video
images images. Fast response liquid crystal
displays would therefore be useful.
Summary of Contributions
• The main findings.
• Be concise/precise! (Eg: EVERY symbol in
every equation should be precisely defined!)
• Do not say “new” or “novel” yourself, which
should s oud be mentioned e to ed by ot others. e s
– Some journals j do not allow such terms at all
• Explain how your experiment is done.
• Show how your equations are derived –
only critical steps should be listed.
• How your sample is characterized.
• You may need to have both of these two
components if you have both theory and
• Identifying all relevant work (with
references) references), even if it is old or does not
solve exactly the same problem.
• Generally, references should be locatable,
avoid use of “personal communications”.
• Do not “bad-mouth” others’ work.
Results and Discussions
• Show your y experimental p results or simulations
• “Figure Figure 1 shows …..” …..
• “It can be seen from figure 1 that ……”
• YYou may want t to t highlight hi hli ht some hidd hidden
results by “It is worth mentioning that …”
• You should compare your results with others.
• This is one of the most important parts.
• A short summary of what has been
reported in this paper (like your abstract).
•But, , do not simply py copy py yyour
• Interpretation of your work: pros & cons cons.
• Limitation of your work work.
• MMaybe b suggestions ti ffor ffuture t work. k
• All sources should be referenced.
• Reference numbers should follow the
sequence in which they are cited cited.
• Different journals have different formats for
references – take note!
• Eg Eg. IEEE: Author Author, Journal Journal, Vol Vol. No No. pages
• Under which grant the project is carried
• Wh Whoever hhas hhelped l d you iin fi finishing i hi thi this
• If the contribution is significant, g yyou
consider to list the person as a co-author.
Choose to cite
• Your reference database can be built up
by reading papers from time to time.
• Or by searching databases, such as “Web
of fS Science”. i ”
Writing gy your paragraph p g p
• An easy way to start a paragraph is “Figure 1
shows …. ”
• Try to avoid very long sentences, i.e.
separate into many short sentences.
– Long g sentences are difficult to comprehend. p
– Purpose p is to let readers know exactly y what yyou
are doing, not to confuse the them!
• The caption of the table should be on top
of the table.
– For a figure, figure it is below the figure figure.
• Different journals have different formats for
• Eg. For APL, JAP, use Table I, Table II,
rather than Table 11, Table 22.
• They should be numbered according to the
sequence they appeared in the text text.
• They should have figure captions to explain
what is in the figure.
• You must refer to EVERY figure in the text.
• Postscript format is generally preferred, Tiff is
not t good d as it iis ttoo bi big and d hhard d tto kknow
whether you have lost some resolution.
HHow tto get t your paper accepted? t d?
– What is the contribution of the paper?
– Does the author explain the significance of this
– Is the paper clearly written and well organized?
– Does the introduction state the purpose of the
– Are the references relevant and complete?
Supply pp y missing g references.
– If the paper is not technically sound, why not?
– If the paper is too long, how can it be
Submitting your paper
• Nowadays Nowadays, online submission is popular or
• Check for example:
Reviewer’s Reviewer s Recommendations
• Paper p Contribution
• Technical Quality
• Originality g y
• Thoroughness of Results
• Clarity of Presentation
• Adequacy of Citations
• Relevance to the Field
• Confidential Comments to the Editor:
• Comments to the Author:
How to write your y draft
• Nowadays – very strict – Template often
•Eg: g IEEE-Trans: Double column format, ,
using IEEE style file (Latex or Word).
• Sequence: title, author, abstract,
keywords, y , main text, , references, , tables
and figures to be cited/placed within text.
• The results of research should be recorded and
maintained for analysis and review, both by
collaborators and other scientists. Exceptions for
preserving privacy, patent etc.
• Misconduct is completely unacceptable
– Fabrication of data ,
– selective reporting of data with the intent to mislead or
– theft of data or research results from others.
Ethics and Responsibilities of
Publication & Authorship Practices
• Th The authors' th ' central t l obligation bli ti is i tto present t a concise, i accurate t account t of f the th
research. A paper should contain sufficient detail and references to permit
others to repeat the work.
• Proper acknowledgment. Authors should cite publications that have been
iinfluential fl ti l in i determining d t i i th the nature t of f the th reported t d work. k Information I f ti obtained bt i d
privately, as in conversation, correspondence, or discussion with third parties,
should not be used or reported without explicit permission from the investigator
with whom the information originated. Information obtained in the course of
confidential services services, such as refereeing manuscripts or grant applications
cannot be used without permission of the author of the work being used.
• Authorship should be limited to those who have made a significant contribution
to the concept, design, execution, or interpretation of the research study. All
those who have made significant contributions should be offered the opportunity
to be listed as authors. Other individuals who have contributed to the study
should be acknowledged, but not identified as authors. The sources of financial
support for the project should be disclosed.
• Plagiarism constitutes unethical scientific behavior and is never acceptable acceptable.
• It is unethical for an author to publish manuscripts describing essentially the
same research in more than one journal of primary publication. Submitting the
same manuscript to more than one journal concurrently is unethical and
• When an error is discovered in a published work, it is the obligation of all
authors to promptly retract the paper or correct the results.
• All collaborators ll b t share h responsibility ibilit ffor any paper th they coauthor. th
• Only persons who have significantly contributed to the research
should be listed as authors authors. The author who submits the paper for
publication should ensure that all appropriate coauthors and no
inappropriate coauthors are included on the paper, and that all
coauthors have seen the final version of the paper and have agreed
to its submission for publication publication.
• Some coauthors have responsibility for the entire paper as an
accurate, verifiable report of the research. These include, for
example, l coauthors th who h are accountable t bl ffor th the iintegrity t it of f th the
critical data reported in the paper, carry out the analysis, write the
manuscript, present major findings at conferences, or provide
scientific leadership for junior colleagues. Other coauthors may have
responsibility ibilit mainly i l for f specific, ifi limited li it d contributions t ib ti tto
• Every coauthor should have the opportunity to review the
manuscript before it is submitted for publication publication. All coauthors have
an obligation to provide prompt retractions or correction of errors in
published works. Any individual unwilling or unable to accept
appropriate responsibility for a paper should not be a coauthor.
• The editor of a journal has complete responsibility and authority to
accept a submitted paper for publication or to reject it. The editor may
confer with associate editors or reviewers for an evaluation to use in
making g this decision.
• An editor should give prompt and unbiased consideration to all
manuscripts offered for publication, judging each on its merits without
regard to race, race gender gender, religious belief, belief ethnic origin origin, citizenship citizenship, or
political philosophy of the authors, and respecting the intellectual
independence of the authors. Situations that may lead to real or
perceived conflicts of interest should be avoided.
• The editor and the editorial staff should not disclose any information
about a manuscript under consideration to anyone other than reviewers
and potential p reviewers. Unpublished p information, , arguments, g , or
interpretations disclosed in a submitted manuscript should not be used
in an editor's own research except with the consent of the author.
• An editor presented with convincing evidence that the substance or
conclusions of a published paper are erroneous should promote the
publication of a correction or retraction.
• Review by y independent p scientists pprovides advice to editors of scientific jjournals
concerning the publication of research results. It is an essential component of the
scientific enterprise, and all scientists have an obligation to participate in the
• Privileged information or ideas obtained through peer review must be kept
confidential and not used for competitive gain. Reviewers must disclose conflicts of
interest resulting from direct competitive, collaborative, or other relationships with
any of the authors, and avoid cases in which such conflicts preclude an objective
• RReviewers i should h ld jjudge d objectively bj ti l th the quality lit of f th the research h reported t d and d respect t
the intellectual independence of the authors. In no case is personal criticism
appropriate. Reviewers should explain and support their judgments in such a way
that editors and authors may understand the basis of their comments.
• Reviewers should point out relevant published work that has not been cited by the
authors. Any statement that an observation, derivation, or argument had been
previously reported should be accompanied by the relevant citation. A reviewer
should also call to the editor's attention any substantial similarity between the
manuscript p under consideration and any y ppublished ppaper p or manuscript p submitted
concurrently to another journal.
• A reviewer should treat a manuscript sent for review as a confidential document. It
should neither be shown to nor discussed with others except, in special cases, to
persons from whom specific advice may be sought; in that event, the identities of
th those consulted lt d should h ld bbe di disclosed l d tto th the editor. dit
• Reviewers should not use or disclose unpublished information, arguments, or
interpretations contained in a manuscript under consideration, except with the
consent of the author.
• Start from the listeners!
• Figure out what they want to know.
• Select the important parts of your paper.
• Be prepared to skip parts of it.
• You may present it in another way or order
compared to the paper paper.
News Headlines Method
• News headlines: right at the start, start present the results of the paper (in the
form of a list with results or orally). Try to make the listeners interested!
• Title: show clearly y what is inside.
• Outline: an overview of the presentation is shown.
• Background and introduction: the audience is introduced to the subject
and the studied problem, e.g., by presenting an example.
• Related work. How others have tried to solve the problem?
• The main part: Method and the results described. Supporting
arguments, investigations and sources described. A few technical details
• Summary: as an ending can you summarize your presentation and
discuss open problems, future work , reading tips, references to
interesting web pages etc.
• Thank you. Any questions?
A few Don’ts
• Don't impress the audience with hairy
theory. y Avoid a lot of technical details.
– Use motivating diagrams/examples instead.
– Avoid to be too wordy.
• Don't present p more material than yyou
to: rather spend more time on one thing.
• Don't hide the problems with your method:
they'll y find out…
Have some notes
• If you y use slides yyou
can write the
supporting notes you need on copies of
the slides. You may y also need some extra
material with additional notes.
•Do not write your presentation word for
word and then read it for the audience –
this gives a dull impression.
• Speak clearly!
• A good way for non-native speaker is to
speak k slowly l l
– Generally one slide one minute.
• Pretend that you are talking to the last
Respect Time Limits!
• Respect the planned time. Clock your
presentation in rehearsal rehearsal.
• Be prepared to remove the less important
slides lid at t the th end, d if you get t lack l k of f time. ti
Decide which before!
– Often at conferences, the chair of the session
will warn you when there is some minutes left,
and d when h your ti time iis up, h he/she / h may even
stop you from finishing! This is not the last
impression you want to give the audience audience.
• Face the audience.
• Do not put your hands into your pockets!
• Do not face the white board or the slides slides.
• Decide in advance if the audience may interrupt
for questions, or if you want to take them at the
• Try to identify the questions to your presentation,
and formulate your answers. Parts of these
answers you may want tt to give i even if not tasked! k d!
• YYou can also l prepare extra t material t i l or slides lid
which answer the questions you anticipate, just
• Give your presentation in advance advance, with or
without listeners, and correct the things
that do not work well well.
– It is i very helpful h l f l ffor beginners. b i
• Clock the presentation.
Stage g fright g
• There is no simple recipe concerning stage
fi fright. htSome S nervousness is i normal land dmeans
that you are sharp.
• A way to master the problem is to be well
prepared prepared. Then you feel safer and this often
reduces the nervousness.
• Also remember that training gives skill! In this
course you get the chance to train in front of
Dealing g with Questions Q
• The self-admirer: "Have you read Dr. Putz'
– The person just want to show his knowledge or
experience to the rest of the audience.
– Compliment p the gguy, y, then move on.
• The attacker: "Your work is boring/irrelevant/
– He wants to get a point or two by showing
weaknesses in your work.
– Be polite, you may say we can discuss this
• Genuine interest
– The most welcome question. Try to answer correctly
and d hhonestly. tl DDon't 't be b afraid f id t to say "I ddon't 't know" k "
when you don't!
The student presentation
To be held here, , Week 13
You will each be required to give a short talk
Th Thursday d 28. 28 AApril il (All part-time students +
additional FT students)
Two other days in Week 13 (to be arranged
between 26 26-29 29 April and announced)
Title of Presentation (think (
Nanyang Technological University
@ t d
Your Presentation – Details:
I will email you all stating on when your
presentation t ti will ill bbe.
You should present no more than 4-5
slides and take no more that 7 minutes
Background / Introduction /
1) What Am I doing? Give an example
2) Why am I doing it?
3) Related work – how have others
tried to solve the problem(s)
4) Method of your work/Diagrams
Main Part and Conclusions
1) Method and Results Results.
2) Supporting theoretical arguments –
why did it work???
3) A few technical details details.
4) Summary, future ideas.