Bhat's Notes

31st January 2020

Coverage by Bhat Dittakavi on “Becoming a holistic Researcher” by B.S.Murty, Director of IIT Hyderabad

“Where you have published matters than how many you have published.” People with european PhDs won’t get hired for teaching in IITs as they don’t do any course work during their PhDs. Goal must be research publication in the top journal.

If the supervisor is not quality conscious, then student is also not quality conscious.

Three qualities of a good researcher

1) Talk with your friend about your research. Research culture is important. We used to teach ourselves every day from 9pm to 10:30pm.

2) You should go to your bed thinking about your problem. PhD student doesn’t have a 9-5 job but she shall pursue her work in the lab for all of her waking hours. Scholarship is not for the job but for the scholar you are.

3) Have a 11th grader excited about your work. You are doing something new and the feeling that you are doing something new has to be there.

What is a good PhD?

The one that makes either scientific or technological impact. PhD produces more questions than answers.

How to make an impact?

1. Proper choice of topics

Once you choose a broad area, you shall come up with at least ten unanswered questions. What they haven’t done is what you shall find from what they did. This is called reading between the lines.

You start like Einstein in year one. In year two, It seems there are others doing something. In year three, It seems others are better than you. In year four and above, It seems you are doing nothing. This is when you are ready to submit your thesis! PhD is like you do more and more about something very less (barrow area) until you learn you know nothing!

Your work must be a ready reference for your juniors.

Should you choose a state of the art or modern area or a topic on a proven fundamental concept? Choose a topic where you can actually make an impact. Hot topics is where you can make impact very quickly. Many a times, in such hot topics, small incremental work becomes very visible. The fundamental understanding may not be much from such work. Is it the new hot topic that you researched? New understanding gets you much more visibility than the new topic alone.

Is the new understanding you bring making an impact or not?

2. Thorough literature research

Two schools of thought. Should you go through the existing literature and get biased or should you not go through the literature so you have unbiased original work? It is important to learn whether someone has already come up with similar ideas and you might be reinventing the wheel.

PhD is a degree that gives you training for you to do research in the future. The faster you do, the better you can take the research journey to the career. Quality is important too. Results from some experiment may warrant you some publication but you shall explain and reason out the results. Failing is not a sin in academics as long as you could explain the results of your experiments. I know an instance where someone got his PhD despite failing. He just made analysis on his findings through his thesis.

One of my close friends from IIT Delhi worked with a guy who made an alloy that becomes glass before it melts. In particular work they got the result and got even a publication in Nature. My friend tried for 4 years and he couldn’t reproduce the result.

There is a journal of irreproducible results in materials!

3. Careful planning and execution

A thesis well planned is a thesis half done. Don’t say first three years none of the experiments succeeded but the results of fourth year succeeded. This happens because of the development of your understanding over those three years. Execution is the easiest and planning is the most important thing.

4. Detailed analysis of the results to bring out new insights

What else is expected from you?

1. Ability to carry out independent research

At some point of time, you will be a manager. In order to guide someone, Ph.D is a training ground. Guide being busy man is a blessing in disguise as it makes you more independent. Once you become a faculty no one guides you but you have to guide the students.

A student who comes up with results is better than the one who obediently asks her supervisor for work. Don’t go to your supervisor before you plan some work.

2. Ability to communicate your results

Seminars and paper help you.

3. Ability to handle a number of experimental facilities

This is domain specific. This leqrning becomes handy for life time.

4. Ability to work in a group

Post doc supervisor seeks this kind of ability from your supervisors recommendation letter. Learn to give and take. Be the one to communicate well in your group. Ask your peers to read your paper and learn from each other’s scrutiny. We can criticize others’ papers well and hence each of you learn. Form your own group so it does 1) fundamental talks as a regular activity. A group of 6 people means you will have to give the talk only once in every six weeks.

5. Ability to supervise others

Post doc becomes very important for this reason. It is expected that guide supervises you in PhD. In postdoc you will learn to be a supervisor of a couple of researchers. That’s the learning ground of supervision.

Summer interns are an opportunity to learn how to supervise.

6. Strong fundamental background

As a researcher you shall read text books. In those days, we used to write a synopsis from the hard copy of the paper we read. Now you do keyword search and download a paper for future reading that never happens.

Researchers knowledge is like a Boltzman’s function that never goes to zero. It peaks but never goes to zero. Just keyword search online may make you miss out on key papers. Identify an important list of journals and fix up a day in the week to go through those journals. Go through the titles of the content page. Once you see a title is interesting to you, download the paper and read it. This is better than just a keyword search. This helps you not o ly become a researcher but a teacher with peripheral knowledge.

How to write a good paper?

1. Problem selection

1. New idea to old problem works.

1. Innovation in experimental or modeling methodology

1. Collaborate whenever possible

Utilize the best possible researchers. I know someone with really high h index. He told his secret. He collaborates with many people. He looks for people who can get him some help. You should know how to write compelling emails.

5. In depth analysis

Don’t rush. Did eep analysis.

6. Publishing your work

Choice of the journal or conference

-Impact factor (what and where people go to find the new works.)

A well respected journal may not make a hihh Impact! Just watch out. Choose the best one. There will be always a journal that accepts your paper. Hence it is in your hands to publish your work in a really good journal. Conference proceedings have a higher value in computer science field whereas for material science publishing in a journal is critical.

-Indian and international journals

At least one paper must be in Indian journal, straight paper written for that Indian journal or conference. We shouldn’t publish in India because your paper for rejected elsewhere,

Quality of paper

Citation index is important. Job role of associate professor or professor at IIT requires many citations. Are your citations high or not?

Plagiarism

Be careful with it. Ph.D can be withdrawn and supervisor might lose his role.

Impact Factor

No of citations in one year divided by no of papers published in previous two years.

Always try to publish in age old journals.

Citation index is important for faculty. Before choose your guide, search for the citations of the guide.

h index is about the consistency and recency. 50 h index means at least 50 papers where each paper has 50 citations.

Citations = Papers = h

This gives visibility in outside world. Outsiders know you only.by your papers and they invite you to be a keynote speaker based on it.

Peer recognition is what ultimately you want. As long as you do good research, you will be recognized.

Please work hard and develop the passion. If you really want to be a researcher, be passionate about it.