We should not judge people by their peak of excellence; but by the distance they have traveled from the point where they started.

Saturday, June 28, 2008

MS vs PHD - The difference.

When I was in India I often used to think what is a PHD? How different it is from a MS? In this post I try to answer some of the basic questions which occur in any students mind about the difference between a PHD and a MS program.

First of all let me give you a brief background on the type of Universities in US.
Different universities could be categorized broadly as.

1. Universities which are research oriented -

The main focus is research. These universities are very well funded. They are
mostly the top universities and very tough to get into.

2. Universities which are research / commercial oriented -

These universities have main focus on research but they also produce candidates
for job market. Like MS students. But 95% people in here are PHD students.

3. Universities which are commercial oriented -
These universities do not do research. Their main aim is to give students degree.
Most of them do not have PHD programs also.

Now I explain the differences between a MS and a PHD.

1. Duration - The first and most important difference is duration. MS duration is fixed. In most of the universities it is upto 2 years. PHD duration can vary from 5 years to more, depending on the professor, university policy and the research that you are doing.

2. Both MS and PHD candidates have the same course work to do for the first two years. So essentially a PHD candidate also gets a MS degree after the completion of the first 2 years. What differs is the rest 3 or more years which he /she spends on doing research.

3. Coursework - First two years are spent taking different courses which helps the student in his future research work. These courses are related to the work he / she wants to do. There is lot of autonomy in terms of choosing the courses. This is a credit based system. The projects which you do in these courses are sort of mini research oriented. So slowly one starts getting the feel of the research oriented environment.

4. Background -
Most of the candidates who appear for PHD have very strong background in research. They would have shown inclination towards research in their previous work. They would have published some papers, have done good work in industry, graduated from a reputed institute. PHD candidates thus show the aptitude for research and getting admit for a PHD program is tougher than getting admit for a MS program. A PHD is sort of commitment from the student. It requires patience, the aptitude towards research.

5. Adviser - Every student works with his adviser. He guides the student initially to identify a problem, find the correct thinking procedure, giving directions, help in giving ideas etc. Slowly as the student matures adviser's involvement grows less. Student has to find a path on his own, identify problems, work on the solutions, fail, recover. The important part is the procedure of thinking, which the student develops ones he goes through this cycle.

What a PHD teaches is how to think independently. How to identify problems, how to seek solutions. May be that is why the PHD in its full form is called "Doctor Of Philosophy". So once student has understood the technique, he concentrates on the research work. The ideas on which he is working is termed as a technical paper and it is published in reputed conferences like say for storage systems "FAST", "USENIX", , "OSDI", "RTSS" etc. There are many such conferences depending on the area you are working in. These conferences attract eminent personalities from academia and industry and they judge the quality of your research work. Links for these could be find in one of my earlier posts. Many students start publishing their work as early as in the first year. Many take 2 years or more to get started. It all depends on your previous background, your advisers interests.

6. Research - The standards of quality are very high. No plagiarism is allowed. The research work mainly could be categorized as

6.1 Survey and analysis of existing work and comparison of them -
consider the analysis of the data access patterns in existing storage servers
and how the performance gets affected. what are the observed trends in
accessibility of data in terms of sequential, randomness etc.

6.2 Design of new ideas -
Design of a new file system.

6.3 Tweaking existing ideas into a different form -

How could an existing file system be used on a different device and how it
would affect the performance.

This is a very gross level division. But mainly the work could be split like this.

7. Funding - PHD is a long term commitment. Mostly in 99% cases students get admits from university if he shows a potential to be a PHD candidate and some professors finds him worthy to be working in his group. Professors control the funding. Professors have control on everything. They design the syllabus, teach, design labs, design exams, homeworks. They receive grants from different sources like industry, Government, national laboratories. The grants in US universities mainly arrive from Governments. National Science Foundation (NSF), Defense Advanced Research Project Agency (DARPA), National laboratories are some of main sources which fund the professors to carry on the research. Some part of it comes from Industry. The professor has to utilize this budget on his students. Thus he sponsors the students and funds their education. He pays their fees. He gives them stipend for monthly expenditure. So it is like a mini job where you have schedules, deadlines and are paid. So a PHD student gets admit only if some professor finds him good enough, so that he could get him in research that he is doing.

The situation varies in case of MS students. MS students do not contribute much in terms of research as their time is spent in doing the coursework for initial 2 years. So many times funding is an issue for MS Students. The available funding to MS students is in terms of Teaching Assistants where you assist professors in conducting labs, grading assignments, solving difficulties of students etc. This is a very good learning experience. This funding comes from the department and not from the professor. Many MS students depending on their potential do get funded by professors and work as research students. It varies from student to student.

8. Many times student switch from MS to PHD program if they find the atmosphere encouraging and feel they could commit to the work. A lot of factors go into taking this decision. Professors always look out for good students and encourage MS students to go for PHD.

This is a different process, than what we find in India. In India only IITs work in this way. But the scale to which this kind of operation is done in IITs is also minimum. MS is a good process to understand all these details. It gives you enough exposure and food for thought, that you can take some decisions about your future and the direction in which you would like to move.

The process makes you more wise and mature. This exposure is what that matters because it makes you aware of so many different perspectives.

No comments:

Post a Comment