Sunday, September 5, 2010

Funding

Q. Will I get funding with an MS admit?
or
Q. How soon can I expect to get funded after joining my Masters program at a US graduate school?
or
Q. How do I go about getting funding as a Masters student?


MS admissions hardly ever come with scholarships or fellowships right with the admission itself (unlike in the case of PhDs). Most MS admission letters would read "The MS is an unfunded program and students should not rely on securing departmental/university funding for attending the graduate program" and or something similar. However, that only means that you are not "assured" to get funding, it doesn't mean that you will not be able to get funding at all.

Sources of funding for Masters students typically include research assistantships, teaching assistantships and if neither of those work out, other jobs like library/administrative/IT assistant, etc. The ease of actually getting funded depends on a lot of factors like the particular university in question, the economic situation, department, etc. Funding will be tougher to get in universities with very large admit pools. E.g., if a usual batch of MS students in a department is 200 or 300-strong, that increases the competition for the limited amount of funding slots. The economic situation (e.g. recession) may affect the amount of funding the university/its professors are getting from the government, defence agencies, industry, etc. for various research projects. Even the department you are admitted into has a part to play. Some departments are in general wealthier (as far as funding is concerned) than others. These are just some of the factors that affect funding, but the best way to get a good picture about the funding scenario at a particular department in a particular university is to talk with a current student there. Forums like edulix.com are good sources for such information. But it would be best to get the contact information of a current student from the university's directory and mail him/her directly to get an accurate picture of the current funding opportunities there.

I can do my part by talking a bit about funding opportunities at Stanford, specifically the Stanford computer science department. The funding scenario for Stanford CS students is quite good. Most people I know managed to get funding at least by their second or third quarter, if not right from the first quarter itself. Of course, it involves a lot of patience to contact different professors, possibly interview with them, etc. and repeat this as many times as necessary. But students who took the funding-hunt seriously did eventually manage to get funded - either as RAs or TAs . CS students also have the advantage of easily applying for RAships in other departments like Psychology, Geophysics, Linguistics, etc., where there may be projects that need students with programming expertise. They thus have plenty of opportunities (because of their programming skills) in places apart from their own department. This is a luxury which students in other departments might not have. Also, from what I knew/heard at Stanford, there certainly were funding opportunities for students in other departments too, but probably not as abundant as those for CS students. It's always better to cross-check this information with current students at the university.

A word about on-campus jobs apart from RAships and TAships. It's not uncommon to be unfunded in the very first quarter/semester. Professors usually want to know you first or evaluate you in a class before they give you an assistantship under them. Thus, with no or little background to show, it can be tough to secure an RAship or TAship right in the first (or even in the first two) semesters. An alternative way to fund yourself is through other on-campus jobs like library assistant, administrative assistant at a particular department, etc. These jobs are usually posted on an internal careers/job listing website and pay on an hourly basis. The salary might not be comparable to what you get as a RA or TA, and you also miss out on the best part of the RA/TA compensation, i.e. tuition waiver. Nevertheless, it can be a source of income to at least offset the day-to-day living costs and saves you from spending that much out of your/parents' pockets. Moreover, this should be a temporary job that you do while you keep your hunt on for TAships and RAships.

My final advice would be that if you are getting into a highly reputed university and into a very in-demand program (where you are assured of getting handsome jobs after graduation), please do not reconsider your decision to join just based on the fear of getting funding. Even if you do not get adequate funding and end up owing a substantial education loan, you should be able to repay that fairly easy with your post-graduation job. Having a Masters degree from a university like Stanford, MIT, Berkeley, CMU or any of the other top US universities is definitely well worth it.

Monday, August 30, 2010

Selecting Universities

Q. How many universities do I apply to?
or
Q. How do I decide which universities to apply to?
or
Q. What are the different things to keep in mind while choosing the universities to apply to?


First, lets talk about how many universities one should apply to. The easy answer is the more you apply to, the better your chances of getting into a university of your choice. Practically however, the number of universities you can apply to is limited by the application costs. This includes the base cost of the application itself which can range anywhere from $45 to well over $100 (depending on the university), and the postal charges in shipping the application material. Depending on your spending budget, you can apply to 8, 10 or more universities, but I would advise applying to at least 8 to have a reasonable chance of making it into a good university.

Once you have an idea of the number of colleges you will be applying to, the next step is to choose them. Some of the factors to consider are:

1. Ranking, general reputation of the college, campus recruitment scenario, average salary of a graduate from that university - these factors are more relevant for MS than PhD
2. Professors, their reputation, research groups, relevancy of research to your own interests - more relevant for PhD than MS
3. Course structure and specializations offered, whether there are enough good courses offered in your area of interest - for instance, there might be good CS universities which might not have enough exciting courses on say, AI or Theoretical Computer Science, since their area of focus is something else. Make sure that the university offers enough courses related to your interests.
4. Flexibility of curriculum, whether it is possible to take interesting courses from other departments as well in addition to your department's core courses - for instance if you are specializing in AI within CS, it might be helpful to take a few courses from other departments like mathematics, statistics, psychology, linguistics, etc. and it would be good to get credits for taking those courses too.
5. Tuition fees and living expenses, i.e. how expensive the education is
6. Funding opportunities

The university website and its students (by getting in touch through email) are good resources to get the above information.

Finally, the last thing to evaluate is how realistic your chances of getting admitted into that university are. Obviously, the higher ranked or reputed the university is, the tougher it is to get into them (with few exceptions). Also, PhD admissions are much more tougher to secure than MS admissions at the same university. There are many online forums (Edulix being one of the most popular ones) where existing students from across different US graduate schools help in evaluating the profile of aspiring graduate students and give them guidance on which universities might be a good bet for their profiles. In addition, you can email students from your interested university and get their opinion on how good your chances are.

The usual algorithm followed is to apply to a few ambitious universities (ones which are out of your reach and would mostly require a miracle for getting accepted there), some gettable ones (i.e. those which should ideally be gettable for your profile, and if luck's on your side, you should get accepted into those) and finally a few safe ones (i.e. those which have been known to give admissions to students with profiles even worse than yours, so you should definitely get an admission there). Suppose your were to apply to 8 universities, depending on your risk taking capacity, you can do a 2-4-2 or 3-3-2 or 4-2-2 split, etc.

Also, there are few universities which have very late deadlines compared to others - these are so late that you can get results from the other universities you have applied to, and still have enough time to apply to these. Keep in mind that such universities do exist, and especially if one of those happens to be a safe or gettable one for you, you can wait for hearing from other more ambitious universities before applying here.

Wednesday, August 25, 2010

Bachelors in X, but applying for MS/PhD in Y

Q. I did my BTech/BE in X field of study. However, for my graduate studies in the US, I want to apply for a graduate programme in a different field Y. What are things to keep in mind for such a situation?

Quite a few students do this every year. Whether this poses a problem at all depends on how close or far apart the two branches are. For instance, it might take a bit of convincing if you are applying for a MS in Computer Science but your undergraduate major is Political Science or even an unrelated engineering branch like Civil engineering. It would be easier if your bachelors was in one of the so-called circuit branches (electrical, electronics, computers, etc.). However, in the end, more than what your bachelor degree says, more important would be factors like your individual courses, projects and work experience, and their relevance to the branch you are applying to.

Here are a few things that can help:
1. Highlight in your SOP any relevant courses done even if it's at an introductory level (e.g. if you are a MSCS aspirant, this might be any CS-related course like intro to algorithms, programming, etc.). Also highlight any projects or work-ex which requires a similar skill set as your intended graduate field of study.
2. If such a relevant project or work-ex exists, get a recommendation letter supporting that.
3. In your SOP, very clearly state the reason or background for considering a branch switch. It should be genuine and logical. For instance, it could be the interest you developed from a course project, or work-ex, etc. which got you really excited in the other branch.
4. Consider giving the subject GRE (in fact this might even be a requirement for some universities when your UG degree is in a completely different field of study - check this individually on the admissions website of each university you are interested in).

Sunday, August 22, 2010

Choosing Recommenders

Q. How do I choose my recommenders?

Choosing recommenders is often a trade off between how famous/reputed or senior the person is and how close the person is to you. Though recommendations from more senior persons (like professors or people with a PhD degree) usually are more impressive (than say from a lecturer who doesn't have a PhD), the closeness of the recommender to you will matter much more. Thus, your recommender could be a highly reputed person in the field you are applying to, but if his recommendation doesn't sound enthusiastic or personal enough (i.e. is too generic), that recommendation will not have a lot of value. It's always better to choose people who have interacted with you or supervised you closely over a considerable period of time, e.g. your final project advisor.

See section 3.5 of DAGAP for a more detailed discussion on this topic.

Statement of Purpose

Q. Could you give me tips on writing a good statement of purpose (SOP)?


You can find tons of resources on the web about writing SOPs and even find a lot of sample SOPs. However, try not to read another person's SOP (or sample SOPs you might find on the web) and derive from that. At least don't look at them before you have written your very first draft by yourself. Looking at someone else's SOP at the very beginning might bias you to write your own SOP in a way, which was probably suited to the other person's profile, but might not work so well in your case.

You should do multiple iterations over the SOP and sometimes it might be necessary to give some idle time between iterations. You can ask peers, seniors and/or professors to review it and give you suggestions on improving your SOP.

I won't talk a lot about SOPs because of the abundance of information already available on the internet on this topic. But do read section 3.4 of DAGAP for further discussion on what a SOP should and should not be.

Some useful links:
http://www.ece.cmu.edu/~mabdelm/statement-of-purpose-tips.html
https://www.cs.umd.edu/grad/writing-statement-of-pupose

GRE AWA

Q. Is the GRE Analytical Writing (AWA) section important?

Yes. Based on my experience as an application reviewer, the importance given to AWA was less than that given to the quantitative section, but definitely more than that given to verbal.

Low AWA scores (<3.5, even 3.5 is borderline) can be a cause for concern.

GRE Preparation

Q. How do I prepare for the GRE?

The GRE general test is a requirement for admission into almost all of the top US universities. It is conducted by ETS and you can find a lot of information about GRE on their website (http://www.ets.org). For an average college student, the quantitative section of GRE (that tests mathematical skills) should be relatively the easier to master. However, the time and effort required in mastering the the Verbal and the Analytical Writing (AWA) sections greatly differ from person to person, depending on one's current proficieny in English. In fact, for most proficient English speakers, the AWA section should not pose much of a problem. However, the Verbal section purely tests knowledge of vocabulary rather than fluency and thus ends up being a hurdle for almost all applicants, since even the most proficient speakers of English wouldn't use most of the "GRE words" in a normal conversation ever in their lifetimes. One usually needs to devote considerable time and effort in getting up to speed with GRE vocabulary. Books published by Barrons, Kaplan and Princeton Review are good resources. So is the web. There are lot of online resources for GRE preparation available these days and even useful discussion forums on social networking sites like Facebook, Orkut, etc. A helpful website for GRE verbal preparation is http://gre-verbal.blogspot.com.

I would highly advise taking a GRE "diagnostic test" - available on ETS website or in the CD that accompanies the Barrons book (or from some other source) at the earliest, typically by end of first year to get a sense of your strengths and weaknesses w.r.t. the different sections and different kind of questions asked in the GRE. You can then direct your preparation accordingly.

DISCLAIMER: This advice is based on what GRE used to be, at the time I gave it (2007). GRE has undergone format changes since then, but hopefully what I said here should still be applicable for the newer formats of the GRE as well.

GRE Score

Q. What's a good GRE score?

The weightage of the GRE score in the overall admission process might differ from university to university. Based on my own experience from being an admissions committee member at Stanford, GRE usually ends up being a filter, such that applicants above a certain cut off are moved forward for further consideration, those getting much lower than the cut off are rejected straightaway and edge cases are at times allowed to be saved if there's something remarkable in rest of the application. In the quantitative section, scores greater than 750 (the closer to 800 the better!) are preferred. In AWA, getting more than 3.5 (or 4 to be on the safer side) should be enough. Admission committees can forgive performance (especially for international students) in the verbal section, but it would still be better to get at least 500.

Again, some universities might be more lenient than the cutoffs I just mentioned, and some might be stricter. With enough practice and preparation, I strongly believe that a score of 1400/1600 (with 800/800 in quant and 600/800 in verbal) and 5.0/6.0 in AWA can definitely be achieved by most students. That should be a safe score to get you past the GRE filter of almost all the top universities.

Reputation of the undergraduate college

Q. How much does the name of the undergraduate college matter?

All top schools do give some weightage to how good the undergraduate college is. When evaluating Indian applications, all top schools have definitely heard of the IITs, but have seldom heard of anything else (even the next-best colleges like NITs, BITS Pilani, Anna Unniversity, etc.). In fact, some times IIITs (Indian/International Institute of Information Technology) get confused with IITs and the IIIT applicants might thus have it lucky thanks to their easily mistakable name. But awareness is on the rise and a lot of the US schools now do recognize IIITs (as being distinct from IITs), BITS Pilani and a few other colleges in addition to the IITs. The importance of the "IIT" brand varies. In some top schools, it may be an absolute necessity to the extent that non-IIT applications are pretty much not evaluated at all. In others, it just adds a bit more of weight and trust to the application. Thus, cases which are on the borderline (say due to borderline GPA or GRE or other factors) and might have been rejected for a non-IITian, might just get accepted because of the "IIT" tag.

Stanford is one of the schools which does give importance to the IIT brand name, but not so much as to overlook other aspects of the application. It thus has an Indian student population such that the number of non-IITians pretty much equals the number of IITians. In summary, if you are at an IIT, stop worrying! If you are not, don't kill yourself about it - there might be a very small number of universities (and this list is becoming smaller every year) that might reject you on account of not being an IITian, but most schools will definitely do a fair evaluation of your application even otherwise. Provided your overall profile stands out, you should still be able be able to into most of the top graduate schools.

MS or PhD?

Q. Should I apply for an MS or a PhD?

This is a very common doubt that a lot of aspirants have. In some cases, the student himself/herself does not have a clue about what he/she wants, in which case they are mostly interested in how this decision (of applying for an MS or a PhD) affects their acceptance chances and funding opportunities. In other cases, the student is quite sure that he/she wants to do a PhD, but is not sure whether directly applying for a PhD right after BTech/BE is the right choice, or if he/she should apply for an MS first and later apply for PhD.

As far as chances of admission at a top school is concerned, a PhD admission is considerably more tougher to secure than an MS admission. The PhD application pool is simply lot more competitive and the admissions committee too has much higher standards of admissions for a PhD admit. If you do not have a very high research orientation and are looking forward to just get an opportunity to study at a top school and/or have the name of that school get attached to your resume, then applying for MS is a much safer bet. Thankfully, a lot of the applications these days have an option that says "consider me for MS if I am rejected from the PhD program". You should definitely select this option and for universities that don't give this option, you would be better off only applying to their MS program.

Of course, things are different if you want research/academia to be your career and are sure about doing a PhD. In that case, it makes more sense to apply for a PhD directly. Such applicants should also keep in mind the things said above about the difficulty of gettign a PhD admit at a top school. However, when PhD is your goal, then the rank of the university doesn't matter as much as the particular professor/research lab you would be working with. Even in universities that might not feature in the top 10 or 20 in most rankings, there might be specific research groups or professors who are highly reputed in their field and do quality research work. As long as you get an opportunity to work with and be advised by such a professor, you should be fine. However if you really really want to do a PhD from a very top school, then you'd be better off doing an MS first and then applying again (your chances should improve once you have done an MS from a reputed university).

With regards to funding, PhD admissions almost certainly coming with funding in most of the top schools. but MS admissions don't. PhD admits get some sort of scholarship or fellowship with their admit which is good for couple of years (or at least one year), and even after that expires, it's easy for them to get research and teaching assistantships. MS students on the other hand have to solely rely on getting RAships and TAships to fund their studies. The ease of getting assistantships differs from university to university. It might be a good idea to contact a couple of graduate students already studying there (it should be fairly easy to get email addresses of students from the university website) about the funding situation there.

Factors influencing the admissions process

Q. What do the US universities look for in an applicant? 
OR 
Q. What are the different aspects that will be evaluated when you apply for a graduate program in the US?

This post lists the different components that make up the MS application. That should give you a brief idea of the different aspects that will be evaluated while an application is under review. The admissions committee does not get a chance to interview or interact with the applicants in any way, it thus has to use UG grades as a proxy for measuring the applicant's knowledge in his/her field of study. The GRE score is another source of input about the applicant's analytical and quantitative skills. Additionally, they look at the projects, internships or work experience (these would be highlighted in the SOP, LORs and resume) to get a better idea of the applicant's background. The SOP and the LOR also help the committee to know the person better and make an informed decision taking into account a lot of things which would not have been otherwise deducible just from the transcripts or the scores.

I might have been a bit vague here and not gone into much depth talking about the different aspects. That's why I encourage you to read my paper "Demystifying the American Graduate Admissions process" which answers this same question in a lot more detail.

Components of a graduate school application

Q. What are the different components of an MS application?


The different components of an MS application are the scores from ETS - GRE and TOEFL, grades (detailed transcripts) from your UG institution, recommendation letters, statement of purpose and optionally, a resume. You need to have these ready before you can apply. You would then create an application on the institute's admissions website, fill in the required details, upload the relevant materials and follow it up by sending rest of the application materials in hardcopy through snail mail. GRE and TOEFL scores are directly communicated to the universities by ETS. For recommendation letters, you usually give the contact details of your recommenders in the online application and the institute then reaches out to them individually. Official transcripts from your college (in sealed envelopes) need to be sent in hard copy to the university. Your SOP and resume would be uploaded while completing the online form.

This is how it usually works. But of course, be sure to read the instructions of your particular university carefully before blindly following the steps mentioned here.

Things to do / keep in mind during undergrad

Q. I have just finished my schooling and have got admission in college X (or I am currently doing my bachelors at college X). I want pursue an MS in Computer Science (or some other branch) right after finishing my bachelors. Give me a general outline of the things I should keep in mind during my undergraduate (UG) studies.

First of all, timeline-wise, you would be applying in the beginning of your fourth year (7th semester). It is important to get requirements like GRE and TOEFL out of the way by around this time. E.g., if you are beginning your UG studies in 2010, you are looking to start a Masters program in the Fall (autumn) of 2014. For that, you need to get your applications in by Dec 2013/Jan 2014. Also, your admission decision will be made based on your profile at the time of application - that would reflect your UG performance in the first three years. The first 3 years are hence crucial.

If you are certain about doing higher studies in a technical field (i.e. MS as opposed to MBA), then the biggest thing that you should concentrate on is undergraduate academics and projects. Every field has some core subjects (e.g. for Computer Science it would be courses like Algorithms, Operating Systems, etc.), and you should try your best to excel in them. Additionally, you should always strive for projects (even if it is a simple class project) which have something novel or researchy about them rather than copying/reimplementing a project that someone has already done before. Consider each project as an opportunity to add something exciting to your profile. Of course, the amount of time you get while doing class projects is restricted, so you might not be able to achieve this all the time, but it's still a good idea to be on the look out for such opportunities. However, definitely try to enforce this in your pre-final year mini project (if your college has such a thing) and your final year projects, where you can devote a lot more time.

Try to build a good rapport with your professors and lecturers (by exceling in their classes, engaging in discussions, doing projects under their supervision) since you will need recommendations from at least three of them while applying. In the summers after your second and third years, be on the search for internships or summer projects. They could be in academia (institutions like the IITs, NTU or NUS Singapore), research labs (Microsoft Research, ISRO, BARC, DRDO, etc.) or the industry (Microsoft, Google, etc.). It would be good to have at least one of your internships in the first two categories. Projects outside outside usual college academics such as these show that you have are enthusiastic about your field of study and have strived to do much more than your normal curriculum during the course of your UG studies. Also, internships are a good opportunity to get a recommendation from a prestigious institution like IIT or Microsoft Research, which can be very useful in case your own college is not among the really famous ones.

Prepare well for GRE in your third year - though there's an option of giving it multiple times, try to finish it off in just one shot and give your best attempt at it. Depending on your current comfort level with the English language, you may need to start working on your English vocabulary much earlier than just the third year, for others, the preparation might just be a couple of months' time commitment. Also, in your third year, start researching different universities in the US that you would like to apply to. Evaluate your choices based on both (i) how good the university is, especially for the areas of research you are interested in and (ii) how realistic a chance you have of getting into that university, given your profile. Use the help of peers, seniors and online forums for helping you make your choices and finally apply well in time before the application deadline.
And it's as easy as that... :).

Friday, August 20, 2010

Experiences as an admissions committee member

Q. Can you share your experiences as an admissions committee member at Stanford?

Sure! This document is a compilation of my personal experience as an application reviewer (for MS in CS) at the Stanford computer science department. It talks in detail about the various factors that affect the admissions process and also highlights common pitfalls that applicants make while applying to US graduate schools.