Swaroop C H

blog books about contact subscribe

To Masters or not

19 Jul 2009

39 people have asked me "The case for master degrees. Should or Shouldn't ?" This article is for those 39 people.

Well, the correct answer almost always is "It depends."

But let me give a few points to think about. Obviously, I'm answering from the perspective of CompSci students. Students of all disciplines can draw analogies to their respective fields.

Question: Do you want to focus on theory or on practice?

If you picked theory, why aren't you thinking of a PhD? If you picked practice, why aren't you thinking of the actual practice of coding and joining a job? Remember, Software engineering is not the same as Computer Science!

In other words, what are your reasons for doing a Masters? Be specific and clear. List down the pros and cons of doing an M.S. degree.

For example, here are few arguments for not doing a M.S.:

Once you have a pros-and-cons list, it will be far easier to decide what to do. If you are still asking the same question, you might as well ask "Should I learn Java or C++?"

Whether you decide to do a Masters or not, I would recommend keeping two things in mind:

  1. Focus on building up an impressive list of things you've done. Follow the Zen Valedictorian Philosophy.

  2. If you already have a few ideas in mind that you want to achieve, then just go ahead and apply The Pyramid Method.

Thinking from a big picture perspective, perhaps The Real Question is: What do you want to do with your life?

If you don't know the answer, then the answer is:

Fix the lifestyle you want. Then work backwards from there.
-- Cal Newport

Update: See "The obsession of Indians with the MBA degree", a similar discussion at StartupDunia.

Comments

Sriranga says:

Brief and a nice post, but after this I get a feeling that it somewhat suggests you not to do MS. It is as if you are citing why you didn't do MS. It's a good thing actually. Shows you can do a lot of things in life without an MS.

Srikanth Thunga says:

Swaroop, I believe your answer to be a force fitted one based on your own exposure. I would like to give my alternate perspective which hopefully will clarify a little more and make the logic a little more objective. It definitely has a lot of my own bias.

The basic underlying premise in your argument is very economic i.e. If there is no further economic value for my brand/career/money-making-ability, I don't have any reason to study further unless I want to do theory(PhD) for my interest. If you had explained a little bit more on the theory aspect of it, you might have hit the goldmine beauty of higher education.

Your argument as well as my own counter argument are not new. It is just that economic returns and social factors have totally masked the real deal of graduate studies.

Premise for most people:
Bachelors forced me into a vocation. Masters will give me more options but I can probably work it out even otherwise through sheer hardwork. The real premise though has to do more than just options.

I would like to clarify a couple of things through this comment:


The power of ideas and meeting people
One of the most beautiful things of doing masters or graduate studies is enjoying the power of ideas. You are in a university system which is the foundation of a lot of new ideas across diverse fields. Even if I am constrained in a single department like computer science say at one of the IITs, I get exposed to so many beautiful ideas which are particular to the university. For a minute, lets assume that I wanted to make a career doing databases on the network(big table research?). IIT bombay has some of the strongest faculty in databases. The interesting thing is that so many of the concepts in an academic setting are borrowed when graduate students research with multi disciplinary intent and exposure. For example, a lot of network theory etc... has gone into database research and applications which might not have been exposed directly in a non university setting. Masters of absorbing ideas than about creating new ideas. But, it has its own good things. I don't just absorb ideas from my own field. Even the design department is very strong in IITB. There might be a possibility that I might have HCI research output put into databases. The world is the limit. Masters gives one an infinite capability to absorb ideas.
Meeting people is another important thing. Meeting people might be totally virtual in a university setting. Sometimes, people would have read papers from a particular prof so many times that you almost know that prof and so on... You just know people, the industry, the setting etc...The conferences etc... will expose you to more like minded people. I have seen this more prominently in graduate studies but a university gives you the right environment to do it. A lot more doors open easily than compared to bachelors. I might have a more meaningful discussion with a social researcher at NIAS Bangalore with a masters in my field than just a bachelors.
PhD is much more diverse than bachelors though very few people realise it.
In my readings of papers by people in the research arena for more than 20 years, you tend to realise that the best people would have had exposure to multiple fields and more often than not PhD actually diversifies your field than constrains it. There is much more exchange of ideas across fields than specialising. You can see this prominently in social sciences but I am sure that computer science also has it. Since its a very nascent field and heavily biased by the industry for economic value of the Phd, the progress has been slow. A lot of concepts of network theory has come from social sciences, a lot of CS is from maths, electronics etc... Any field in CS has been a culmination of efforts from multiple other subjects. In a few years, CS will play the bearer for many fields. Probably, there will be a lot of social sciences research based on CS. For example, fractals is being regularly used in sociology, behavioral finance and the best people in a decade for those fields will be computer scientists as they have the right exposure now to the relevant ideas.


You can sense this opinion in people doing graduate studies either in european schools like max planck or by talking to non Indians in US schools. A lot of bias is especially because of the H1B system of 20000 visas dedicated to the US masters students. The H1B visa especially masks the real value of the decision as it makes the decision as a very economic one.

At the end of it, if you are planning to do a masters for money, go do MS in US. It has not other logic other than pure economic one. If you love to absorb, discuss ideas, do masters wherever you want to do in whichever field is most interesting to you. It will be a beautiful experiences. You will definitely learn new things from other fields too. You have to be open to learning, see and hear from other fields. If you want to not just absorb but create and share ideas with the world, you will have to do graduate studies and like to be linked to an university system for life.

A really stubborn person might say that I might get to do all this even without this environment. The real adrenaline rush of ideas and the power of ideas is there only in the right environment as people who have worked have also experienced sometimes. The environment makes it easier to move much much further in much less amount of time than if you have to fight a lone battle. Startups have been successful from villages too but a tech startup has more probability of success in Bangalore than in Shimoga.

--

On a slightly nostalgic note... Our batch(not just CS) in PESIT was truly exceptional. I don't think we will ever find another batch of that size of exceptional people coming together at one time. The reasons were surely a mix of a lot of things which included location, growth of computing industry in Bangalore & the odd strictness which had its own good perks. Even then, it lacked the power of ideas because of the nature of bachelors education as well as the lack of research culture in undergrad colleges of Bangalore. I am assuming that this factor has a major role in your bias for this blogpost.

--

This comment doesn't do justice for what I want to convey but I wanted the readers of this blog to get an alternate perspective.

My comment is irrelevant if all those questions to you were for doing something different after 2 years thinking they might like the new work more. Knowing people, they probably don't have a clue other than peer pressure as to why they want to do masters. Hopefully this comment will change the perspective of one or two of your readers who might have had a different reason for this question.

Srichand Pendyala says:

Some background: I graduated with a Masters degree in CS from MichiganTech a week ago and I've started working full time since then. I graduated with a BE from VTU (MSRIT) two years ago.

I had this exact same decision to make about two and a half years ago along with a couple of friends of mine. Over the course of the two years that I've been studying, my personal justification for getting a Masters degree have certainly evolved. But at no point did I ever regret my decision, nor did I want to drop out or such like.

I am certainly interested in research (I suspect this is what you mean by "theory", but there's nothing to say practical work isn't research) and I certainly was interested in research two years ago. I did want to apply for a Ph.D, but I chose not to, for the following reasons:


The only flavor of research I had had, was at a small research "lab" at MSRIT. I really had no clue of how real research happens at the bigger places with big labs and groups. And effectively, I had no clue if I was going to enjoy or survive what "real" research takes.
Again, with undergrad, I had spent maybe two semesters attempting to "do" research. I had no way of knowing if I would survive the 3-5 years it normally takes for a Ph.D
With the financial investment that graduate school demands (in terms of lost pay over two years + tuition fees + living expenses), I could not justify an experiment to see if I could like and fit into an appropriate Ph.D program. That said, I was fortunate enough to not have to pay for tuition for my entire graduate program.
Finally, it would have been impossible for me to guess if I would get along with my advisor-to-be and spend a good 5 years under their tutelage successfully. (If you must know, the success rate for Ph.D graduation is drastically lower for those joining with a bachelors degree than with a Masters degree. I'm not sure of exact numbers, but I believe its 30% for BE/BS and about 78% for MS).


Now that I've graduated with my Masters degree with a thesis (Sketch recognition through shaped interaction, HCI), here's my thoughts:


The last two years were the hardest I've ever experienced in my short life. From experiencing -40 C winters, to 70 hours+ of work each week (research + teaching + course work + my own projects), to falling sick and staying in hospital for 5 days (running up a $14k bill), to finally completing my thesis last month, this has been the wildest ride ever. But you know what? I loved every day of it.
Instead of taking a course-work option, I took the thesis option, and of course that meant working on original research. Fortunately, I found a gem of an adviser and things worked out fine.
The day I was leaving my little school town for Chicago IL, I told my adviser that I'd like to get a Ph.D. Having seen a lot of my friend work through a Ph.D program, I'm confident that its what I want. Getting a Masters degree meant that I could experiment with a bunch of areas and choose what I liked after working on it for sometime.
I got into my Masters program fully expecting to work on Machine learning algorithms. My graduating thesis is in Human-Computer Interaction, with little ML involved. Your point about knowing what to do in life right after a BE is totally off the mark. I think my idea of what I'm doing/want to do in life has evolved tremendously as I've learnt more. This may not necessarily be true for everyone, but its certainly my mantra.
This week, I started work at an options trading firm in Chicago, where we use algorithms and models to model market data and help our floor traders make trading decisions. Totally unrelated to HCI, but I love doing what I'm doing (I interned here last summer too). I'm convinced that "area of interest" is a very loose and variable term. I find lots of things interesting and I spend a lot of time with them too.
Finally, I intend to work for a year or two to pay off my debts (mostly from the hospital stay, fortunately I didn't have any tuition fees to pay up) and then get back to grad school for a Ph.D. I have no clue which university I intend to go to nor do I know for sure what I want to work on. It could very well be HCI or Mathematical/Computational Finance or heck, even Computational Physics. But whatever I choose, I am certain that I will go to school knowing what to expect and knowing what the program expects of me. And this, I will attribute to my Masters degree.
Regarding a previous comment about MS in the US not having any logic other than an economic one: I think that its a blanket statement, with no regards to school, program, research or adviser. I think any school in the world is a great place to do research and get a graduate degree, as long as you have an adviser you want to work with, research opportunities that you need (research group, hardware ,labs, grants ... ).


The United States is in my limited experience, the best place to get a graduate degree. I base this solely on the respect for academics and research I see here. Advanced education and research are a serious business here, unlike back in India. My "standing" in the academic community is not based on my color, sex, race or religion, but entirely on my work and contribution to the committee. (I'm sure there are exceptions, but they're rare). This of course means that the best and the brightest from all over the world come to study and work here. Which is the greatest thing that can happen to anyone interested in research.

Example: In my team at work, we have 3 Americans, one Brazilian, Irishman and Indian (me) and two Chinese-origin developers. All with advanced degrees from the united states in different areas. How's that for diversity?

To conclude, I must insist that graduate school in the united states (but not necessarily) is an experience everyone in life must go through. At the very least, it builds character. You really won't know your limits until you push right to them. Here, you get an opportunity to do just that. At anything other than the very least, you'll have an experience of a lifetime, meet like-minded passionate people in your area of interest, travel and meet people of different cultures and background.

Swaroop says:

@Srikanth : Well, I am writing my thoughts here, so obviously it will be based on my own exposure. That is why excellent comments such as yours make a blog interesting :)

I agree with your observation. I am focusing on the value of the education itself whereas you are focusing on the value of college life. I think they are two different things. But choosing to go for a Masters degree for the value of college life is a perfectly good reason. But the problem is that most people who ask the question, in my opinion, have not-so-admirable motivations such as "I am not getting a job, so maybe I should do masters."

Also, I have to clarify that nowhere I said that Theory is a bad thing. Hell no. My point was that you have to realize your inclination to one or the other. Very few people can straddle both worlds.

@Srichand : Thank you for the fantastic thoughts.

Regarding Point 4, I think it deserves more explanation. I think motivations to do something should be clear, even if you don't have the big picture in clarity. Course correction is always required for any plan. Start off in a direction, if it is not working out, then change direction. But if you never start having a direction, then you'll be just swaying whichever way the wind takes you, which I don't think is a great method in the long run.

Abdullah says:

Hi,
I think it is a good initiative from your side to help people seeking advice on whether to do masters or not. But, however, I think that question should be best answered by someone who has studied master and knows what it really is like. If we are speaking of computer science then most of the undergraduate programs offer very little exposure to the core of the subject and hence any plans you make are bound to be according to what you know which again is limited in scope. I mean you never really get a chance to broaden your horizon in UG in India (not sure how it is elsewhere) and Masters is the best option. India has some excellent faculty and excellent universities but they are invisible to the undergraduates.
Regarding your statement about point 4 to Srichand, can you not see that he
has stated clearly that he got into his Masters program fully expecting to work on Machine learning algorithms. But that was what he knew then. These kind of things look really cool when you are in UG but when you really get the exposure, you realize that there are many fields that are much more interesting and far more worth the time then the one you started out with. In fact you would never have
known that such fields even existed.
The bottom line is this. In case you meant that people should have a clear idea as to what they want to achieve via a Masters degree, then, what is clear to one person may not be clear for another. For instance, "I want to do MS to get a good insight into the field and get in touch with people of similar interes" is pretty clear to me. But that may not be so with the people who are working. They will say "OK but what will you do with that?". Does everything you study have to be used? For economic or other things? Is there no worth in developing one's intellect and personality and character? If everyone become like that then that would be a loss to the complete society and we would not have so many of the good things we enjoy today.

Swaroop says:

@Abdullah Why should this question be answered only by those who have done Masters and not by someone who has evaluated Masters as an option?

Srichand clearly knew why he was getting into Masters, irrespective of whether he changed direction or not (which is often the reality)

Would you approve if a person spends 20 lakh rupees or so to do a Masters just because everybody else is doing it?

I hope you see now why I'm suggesting that people clarify their inclinations before taking this decision.

Abdullah says:

I am sorry!!! My comment was not for those who want to invest huge amounts of money in studying abroad but rather for those people who do it out of interest. For those who are really interested, institutions like IISc, IIt are excellent and the tuition fees if any is negligible. Rather they have grants for graduate students. Also, going directly into phd is not the best option as you may not know which field suits you best , you do not know exactly what phd is all about, you do not know the professors, the professors may not know you and so on. The direct successes from ug to phd are very rare as someone already commented.

Also, tell us about how you evaluated Masters option? What were your conclusions?

Sandeep says:

Really useful post, a must-read for those who are in their final year of engineering!
Thanks Swaroop for writing this post and Sriranga, Srichand for discussing and giving more insights to masters studies.

Neel says:

I would like to second Srichand's observations here - graduate school does build character (of course, it is not the only thing that does, consider the army if your tastes are more extreme on the character-building front). The rest of my comment may come off as off-topic, but hopefully sufficiently within context to be of interest.

I am pursuing a PhD at the Institute of Mathematical Sciences (matscience for short) in Chennai, and I joined this program after my BSc at Mount Carmel College in Bangalore. As for the choice between harsh winters and harsh summers, I chose the latter because I felt under-prepared for a long stint abroad. I discarded the option of doing a Masters' because I was reasonably sure that I would eventually do a PhD - I do not intend to go too far from the business of academics and/or research, and a PhD would be considered a pre-requisite. (I save time by doing it after my BSc rather than after a MSc.)

A PhD is a good thing to do even if you don't intend to stay in the usual academic circles for life - usually people shy away because of the teaching duties that come for free with such a career. However, but Google/Yahoo/IBM/et al have excellent research labs which are great places to go to after a PhD (and I gather that one would feel more confident in the research lab environment post-PhD/MSc rather than post-BTech). This sort of a job is attractive for various reasons, and if you have the slightest inclination towards learning and discovering new things all the time, then it's good to know that you can get paid for it too.

Consider music - you can learn to play by ear, with minimal "formal training", by watching others and learning how to read sheet music with some basic intuition about rhythm an scales. There are some, however, who yearn to know what really makes the whole thing tick, and they feel restless (in spite of a certain degree of mastery in practice) about the way they do things. An investment of a couple of years in learning music theory makes some people feel a lot better about how they play music. And there are some who actually don't care - that's perfectly fine as long as you are confident that you indeed don't care.

Do a Masters' if you can actually sense some dissatisfaction with your present status of learning. You may know the best algorithms for a large set of problems, and perhaps the implementation is at your fingertips. But if it interests you to know (for example) that there is a formal classification of problems based on their complexity, and that there are good (formal) reasons for saying that you don't expect to get an answer to the TSP in polynomial time, then you should consider spending some time on discovering more magic than what you've seen so far. Also, the new magic helps you build a strong sense of intuition, using which you can stare at a problem statement and remark that "Oh, don't think you can crack that in linear time, because...".

Having said that - if the world that you are hitherto familiar with comes across as magical enough and you aren't hungry for more, you may find the masters' experience merely frustrating, and you'll spend your time itching to get out of it into the real world where you can write/improvise/hack away at code that you knew to write months ago.

Also, much can be achieved by self-learning, and open courseware programs (such as the one offered by MIT) make it completely feasible for you to gain the experience that you would with a Masters' program at your own pace. If there are no reasons for having the degree itself, then picking up things on your own is a serious alternative. Being at an institute has the advantage of greater levels of discipline and interaction, but the former may be precisely what you want to avoid, and the latter can be made up for in many ways.

A final note on the financial aspect of things - a PhD program (at least the ones I know of in India) comes with a stipend that is getting more and more competitive with time. Funds are being poured into the research industry by the governement and industry alike - and the competition/glamour hasn't really caught on proportionately. Thus if you have somewhat decent abilities, you stand a good chance, and you can be further sponsored in many attractive ways if you are better than decent. Fact that someone like me managed to (at least so far) survive the program should give anybody a lot of confidence :)

Parag Shah says:

Very nice post. I think most students are confronted with this question. I did a masters in CS a decade back, so here are my 2 paise:

To start with I am very glad that I did the masters. I do not think doing a job for those 2 years would have taught me more than what I learned in school. Having said this, I believe I had a good experience doing my masters because I had the opportunity to do research with a very nice advisor and had excellent peers to interact with. A job would never have given me (agreeing with @Sreekanth here) the kind of exposure to ideas and people that I got while at the university. I don't think I learned much in the classroom, but I did learn a lot writing actual code and interacting with other members of the research team.

Beyond the usual learning, writing my theses helped me improve my writing as well as communication skills significantly. In fact it also got me interested in writing to the extent that I started my own newsletter and later on my blog. Now I am not suggesting that a masters be done to inculcate an interest in writing. I am simply stating this as one more benefit I got from my masters.

Beyond just the money making capabilities, a (good) university environment offers exposure to people, ideas, and a certain environment which a job may not offer. The friendships I made back then remain so till date, and many of the things I learned in those early years still influence my thought process in a positive way.

Having said this I would also like to add that not everyone may have the resources and opportunity to pursue a masters education. I do not think all is lost. The online world in this day and age allows us (through some creativity) to simulate a large part of the a creative environment with ideas, and interactions that exist in a university.

To summarize, I think a masters is useful, but only if the university offers the kind of environment which is truly invigorating. Doing an MS just for the degree may not be very useful. In that case it might be better to seek a job instead.

sashank says:

I have gone through this question of whether to do MS/Mtech or not at multiple phases of my life so far and each time my thought process was different .


All through my college , i was dreaming of doing MTech at IITs , despite my best efforts i got disqualified at GATE , got depressed for few weeks , then went out to search for a job , afterall BTech or MTech is just get a job ! ( was my thought that time )
After joining a Govt RnD , i was blown away with the taste of actual research and my thought process only got better after my first date with FOSS software , during this time few of my friends went for part time MTech in BMS and few others MS in BITs Pilani , I did not care for , because iam already having the adrenalin rush due to making things work , exploring how things work and developing new cool stuff , Why boring Studies again ? Why boring exams, preparations etc ? when iam having the fun i want ( was my thought process this time )
And now after 6 yrs of Software engg life , after few unsuccessful stints with few product ideas and entrepreneurship , after getting bored of watching few products which donot add any value to life , i feel there is no proper research went in making this products , now i feel like doing PhD , for exploring , researching to create products of value , which has impact on our lives , which can make this world better place to live . ( this is my current thought process ! )

Swaroop says:

@Abdullah If people are doing M.S. out of their interest in studies, this question doesn't even arise for them :-)
My evaluation and conclusions are really specific to me, and it won't be useful for others (besides, I don't even remember what I had written down)

@Neel Thanks for your thoughts. It's good to see the clarity that you knew that you would want to do a Ph.D someday anyway. That's the kind of thing that I hope students start off with, even if what happens once they reach college is opposite to what they imagined. And thanks for mentioning what students should expect.

@Parag The divide that you mentioning is the same software engineering vs. computer science divide that I referred to originally. The kind of learnings you would get in the two realms will tend to be the learnings that you would expect in a good job vs. a good university.

@Sashank I can really relate to what you are saying, although I didn't have a taste of research and after all my brushes with that realm, I think I'm better off in this side of the world, heh.

@Everyone, Thank you all for the quality conversations. It has certainly been enlightening, and I think the discussion here added more value for the reader than my actual post. Thanks!

Lost in thoughts « *.rng = *.Sriranga Chidambara J L says:

[...] The period of time just after one graduates from College is the most challenging if you ask for my opinion. Everyone is confused. Some of my friends have gone abroad to do MS;  some of them to pursue their academic interest, and some of them coz they couldn’t  manage to get a job they wanted. For a good read about whether to do or not to do MS, read Swaroop CH’s post here. [...]

Feedback

There's no comment box, but please do email me or tweet me your thoughts and criticisms, and I will publish the relevant ones here.