You are on page 1of 47

Research and Publication Ethics-

Understanding Research Process

Dr. P. Arulmozhivarman
Professor & Dean Academic Research
CCE
Vellore Institute of Technology
Vellore
Email: parulmozhivarman@vit.ac.in

3/10/2023 Dr.P.Arulmozhivarman 1
New World Culture
1. Culture of quality
2. World of competition

3. World of information & transparency


4. World of Accountability
( Evidence-Based & Sound Decision Making)
5. World of Social Justice
6. World of Humanity
3/10/2023 Dr.P.Arulmozhivarman 2
TEN COMMANDS
BE NOT LED BY
• REPORT
• TRADITION
• HEARSAY
• AUTHORITY OF TEXTS
• MERE LOGIC
• INFERENCE
• CONSIDERING APPEARANCE
• AGREEMENT WITH A CONSIDERED AND
APPROVED
• SEEMING POSSIBILITY THEORY
• THE IDEA “ THIS IS OUR
3/10/2023
TEACHER”
Dr.P.Arulmozhivarman 3
Daily Life of a Researcher

Reading
Thinking
Presenting

Discussing
Writing
Simulations
3/10/2023 Dr.P.Arulmozhivarman 4
My View

 There is no easy formula for research


and publishing journal papers.
 Keys for good research:

- Work hard!
- Work smart!
- and a little bit of luck!

3/10/2023 Dr.P.Arulmozhivarman 5
Two Concrete Suggestions

3/10/2023 Dr.P.Arulmozhivarman 6
Outline

 How do I find a research topic?


 Getting new ideas and developing results
 Writing a journal paper
 Other issues

3/10/2023 Dr.P.Arulmozhivarman 7
How Do I Find a Research Topic?

3/10/2023 Dr.P.Arulmozhivarman 8
Research Topics

 Should be simple to state


 Not obvious how to do it
 Clear benefit
 Can be broken into steps
 Progress and solution is testable

3/10/2023 Dr.P.Arulmozhivarman 9
How Do I Find a Research Topic?

From advisors, students, collaborators


Brainstorming with colleagues
Review papers, listen to research talks
Teach a course/Give a talk: forced to understand the details
and think hard to prepare for tough questions
Hot emerging fields that could lead to many publications or
easier funding

3/10/2023 Dr.P.Arulmozhivarman 10
Good Research Topics

Should be of interest to both you and your advisor


The work could lead to a well defined set of results
Often can be formulated as a constrained optimization
problem
Timely, useful
Work on the significant parts, ask:
If you are successful, so what?

3/10/2023 Dr.P.Arulmozhivarman 11
Getting New Ideas and
Developing Results

3/10/2023 Dr.P.Arulmozhivarman 12
Starting

 Starting is most difficult. A lot of frustrations.

 Literature search:
* know state-of-the-art
* read a lot
* read selectively

 Discuss with advisors and colleagues

3/10/2023 Dr.P.Arulmozhivarman 13
How to construct the search concept?

Identify keywords List synonyms


& main concepts for each concept

Think of the broader terms


for each concept to get
a better perspective
on your research topic
3/10/2023 Dr.P.Arulmozhivarman 14
When Reading a Paper

 Ask:
- What is the historical background of the paper?
- What is the motivation of the paper?
- What are the contributions of the paper?
- What else can I do?
- Can I do better?

 Get more info from the references.

3/10/2023 Dr.P.Arulmozhivarman 15
Do Something!
 Don’t just keep reading paper forever!

 Try to do something – learn from failures.


Implementing state-of-the-art techniques
to gain insights.

 Learn as much as you could, do as much as you


could. If you don’t do it, you never get there!

 Give seminars, force you to think deeply and


prepare for tough questions.

3/10/2023 Dr.P.Arulmozhivarman 16
Getting New Ideas (1/2)

Brainstorming with your colleagues


Consider different constraints from new technologies
or applications
Combining knowledge from different fields
It is o.k. to start from something small. A small idea
may grow big.

3/10/2023 Dr.P.Arulmozhivarman 17
Getting New Ideas (2/2)

Make sure it is really a new idea. Talk to experts


early, don’t just depend on your own thinking.
Make sure your assumptions and understandings
are correct
Analyze and develop theoretical support
Generalization/Abstraction

3/10/2023 Dr.P.Arulmozhivarman 18
Collaborations (1/2)

More success comes from working with others:


Discuss your idea with your advisor and peers
- discussing help clarify what your idea really is
- get constructive feedback
Regular schedule of meetings
Carefully consider criticism from others

3/10/2023 Dr.P.Arulmozhivarman 19
Collaborations (2/2)

Collaborate with experts, learn from experts.


Collaborations are not easy. Different personalities,
different styles, different opinions, different ways of
doing things.
Don’t worry too much about credits. You establish
yourself through your long term contributions.

3/10/2023 Dr.P.Arulmozhivarman 20
Simulations
Ideas need to be simulated, cannot just depend on
intuition
Make sure all the assumptions are valid and sound
Get quantitative comparisons
Make sure the comparisons are fair
Interpret simulation results. Every behavior needs to be
explained. This often leads to extra insights and new
ideas.
If something work, there must be a reason behind.
If something doesn’t work, there must be a reason
behind.
3/10/2023 Dr.P.Arulmozhivarman 21
Keep Documenting

 Keep notes about your work regularly. Write up each


possible piece of the work for publication.

 Documenting is not writing final journal papers. Don’t


worry about polishing the wording.

 Keep documenting could minimize the pain of


writing the final paper.

3/10/2023 Dr.P.Arulmozhivarman 22
Writing A Journal Paper

3/10/2023 Dr.P.Arulmozhivarman 23
Presenting Research Results

• A paper is to present NEW and BETTER techniques


for solving a PROBLEM, so, … always keep in mind:

- What is the PROBLEM?


- What is NEW?
- Why BETTER?
- HOW MUCH BETTER?

• Emphasize your salient points, summarize YOUR


contributions

3/10/2023 Dr.P.Arulmozhivarman 24
Technical Writing

 Good writing is essential !!!

 Take a technical writing course

 Read a lot: Learn how to tell a good story;


Learn how to write a good paper

3/10/2023 Dr.P.Arulmozhivarman 25
Writing a Journal Paper (1/6)

 Logical presentation with good English


- use top-down organization, outline, logic flow, …
- use other good papers as “samples”

 Clearly separate previous works and your own


contributions. Focus on your own contributions.

 Focus on the novel parts

 Be concise, to the point

3/10/2023 Dr.P.Arulmozhivarman 26
Writing a Journal Paper (2/6)

You know your work, but the reviewers and the


readers may not!
 Reviewers/Readers: in the same broad area, but
may not have worked on your specific problems
 Precise and clear definition of terms
 Motivations and rationales
 Pose potential questions and answer them, give
intuitions

3/10/2023 Dr.P.Arulmozhivarman 27
Writing a Journal Paper (3/6)
A reviewer’s check list (partial):
 Does the paper introduce a new problem or
provide a new solution to an existing one?
 What is the main result of the paper?
 Is the result significant?
Is the paper technically sound (assumptions? correct?
…) ?
 Does the paper provide an assessment of the strengths
and limitations of the techniques/results?
 Is the paper clearly written?
 Does the paper reference appropriate related work?
3/10/2023 Dr.P.Arulmozhivarman 28
Writing a Journal Paper (4/6)

 Any not-so-obvious statement should be supported by


references or simulation results
 Avoid using many adjectives (how large is “very
large”?)

 Give right level of details

 Spell out your points, don’t let your reader guess

 Always check spelling. Use “cheap” English proof-


reader.

3/10/2023 Dr.P.Arulmozhivarman 29
Writing a Journal Paper (5/6)

n Pay attention to packaging. Use block-diagrams,


equations, figures, pictures, tables, simulation
curves, …

Class Symbols Bits


Class I P0, P1, P2, L0, L1, S1, S2 30 bits
Class II L2, L3, GB1, GA1, GB2, GA2 24 bits
Class III C1, C2 26 bits

3/10/2023 Dr.P.Arulmozhivarman 30
Writing a Journal Paper (6/6)
 Revise, revise, and revise. Imaging you are a reviewer
trying to critically criticize the paper. Read it word by
word.

 To get a well written paper, early drafts may often


be completely re-written! Revise carefully > 3 times,
total > 5 times.

 Get others to help. It is hard to find own mistakes.

 Publish fast before it becomes obsolete. Research is


never ending. In many cases, writing a paper when you
have enough contributions, don’t need to wait when the
solution
3/10/2023 is perfect. Dr.P.Arulmozhivarman 31
Template - Organizations

n Abstract
n Introduction
- background and previous work
n Your work
- may split into multiple sections
n Experimental results, comparisons and
interpretations
n Conclusion
n (Acknowledgements)
n References
n (Appendix)
3/10/2023 Dr.P.Arulmozhivarman 32
Template - Abstract
n Purpose: Give overall picture, entice
reader
n Do:
 Be brief
 Give relevant high-level descriptions
 Indicate why your work is new, better
n Don’t:
 Repeat introduction or conclusions
 Give details

3/10/2023 Dr.P.Arulmozhivarman 33
Template - Introduction
n Motivation of your work: problems to solve, why they
are important, why they are hard

n Review previous works (don’t just give references, give


one or two sentences summarizing the key points),
what are their problems you will address

n Explain why your work is different and better:


very briefly summarize your new ideas, give intuitions
why they are good

n Summarize your contributions

n Provide a roadmap to your paper:


3/10/2023 Dr.P.Arulmozhivarman 34
organization of rest of the paper
Template - Your Work
Clean problem formulations, clean notations, clean
solutions.

Focus on your own work, don’t include background


material or other people’s work here

Try to be as general as possible

Limit the complexity: don’t put up things unrelated


to the main points of your contributions

More significant part first

Put less important details (e.g. proof of a theorem)


into appendix
3/10/2023 Dr.P.Arulmozhivarman 35
Template - Experimental
Results
Clearly state the complete simulation
conditions so that others can compare your
work

Need quantitative performance evaluation

Don’t just describe, interpret

Elaborate effects of assumptions

Any magic number needs to be justified

Draw conclusions
3/10/2023 Dr.P.Arulmozhivarman 36
Template - Conclusions

Repeat what is great about your work


Summarize your contributions
Mention how general the work is
Hint on future work

3/10/2023 Dr.P.Arulmozhivarman 37
Template - References

Include references for mentioned previous


work

Be careful about the correct format

3/10/2023 Dr.P.Arulmozhivarman 38
Paper Review (1/2)

About 3-4 months after acknowledgement from EIC, if you don’t


receive review results, check with the AE politely.

After receiving the review: fix the problems if you agree with the
reviewers’ comments.

Reviewers are experts. Even if the you don’t agree with the
reviewers’ comments, those points represent parts that you can
improve (make it more clear).

Reviewers are not always correct. It is o.k. to disagree with the


reviewers and clarify the misunderstandings.

Write detailed “response to the reviewers”.


Address every point raised. Try to be constructive and positive.
3/10/2023 Dr.P.Arulmozhivarman 39
Paper Review (2/2)

• No matter how harsh the reviewers’ comments,


don’t get frustrated. Don’t use emotional words..
• Be patient. The paper may go through three or
four review cycles.
• Thank the reviewers.
• Resubmit in two months.

3/10/2023 Dr.P.Arulmozhivarman 40
Other Issues

Rebuttal and Appeal process

Be aware of overlength page charge

Conference papers vs. journal papers

Double submission

3/10/2023 Dr.P.Arulmozhivarman 41
Writing Papers, the Good
Construct a careful argument
– Structure design and implementation sections like a
logical proof
– Known: A, B, C; A+B  D, D+CE. Hence: E
– Outline argument in introduction, add transitions and
back references to previous arguments
Identify your assumptions
– Crucial to readers’ understanding of argument
– Clearly define them, justify via citation or brief
explanation
– Understand the audience, which assumptions can be
made implicit, whichDr.P.Arulmozhivarman
3/10/2023 cannot 42
Writing Papers, the Bad
Present a system approach without
context, justification, or differentiating
from related work
Present a system and lots of analysis
numbers without clearly stating the
argument
Piles of math formulas or figures without
high level explanations of their meaning
Compare against bad implementations or
under unfair conditions/assumptions
3/10/2023 Dr.P.Arulmozhivarman 43
Summary
 Imagination, Composition skills, Hardwork, Passion are
essential
 Finding a research topic?
- discuss with collaborators, review papers, teach
courses, work on emerging fields, …
 Getting new ideas and developing results
- search literature, implement existing techniques,
brainstorm with colleagues, give seminars,
combine different fields, do simulations, …
 Writing a journal paper
- articulate your contributions, good writing, take a
technical writing course, …
 Other issues
3/10/2023 Dr.P.Arulmozhivarman 44
Conclusions

3/10/2023 Dr.P.Arulmozhivarman 45
References
1..R. Kothari, Research Methodology
Methods and Techniques, 2/e, Vishwa
Prakashan, 2006.
2. Bendat and Piersol, Random data:
Analysis and Measurement Procedures,
Wiley Interscience, 2001.
3. Donald R. Cooper, Pamela S.
Schindler, Business Research Methods,
8/e, Tata McGraw-Hill Co. Ltd., 2006.
Dr.P.Arulmozhivarman 46
Thank You

3/10/2023 Dr.P.Arulmozhivarman

You might also like