Chance of getting into a top school w/o publications (Fall 2012)

[josh-cs]

Hi all,

I graduated with a double B.S. in CS and Math last year. I've worked full time as a Software Engineer for over 2 years now. I'm starting to think that being a code monkey just isn't what I want, but I love CS and I'm really interested in research. I did excellent as an undergrad (almost a 4.0), and received many merit-based awards, but it was a pretty small/unknown school. I know I can get some great LORs and I'm confident that with enough prep I can get at least a 1400 on the GRE.

My question is, do I have a reasonable chance at getting into a top school with no publications or real research experience? I have a few months now to prepare if I want to apply for Fall 2012, but my time is extremely limited (I work up to 70 hours a week) and I want to have some idea of what my chances are. Also, would 2+ years experience in the CS industry help my chances much?

Getting in a Big-4 schools would be a dream, but from what I've read, I wouldn't have a chance in hell w/o publications. Other than those, I am most interested in:

Princeton

Harvard

UNC Chapel Hill

Duke

Georgia Tech

Yale

UT-Austin

Johns Hopkins

Also, ML and Software Engineering are what I am most interested in studying.

Thanks in advance!

Edit: s/UT-Texas/UT-Austin/

Edited July 13, 2011 by josh-cs

[stackoverflow]

Hi all,

I graduated with a double B.S. in CS and Math last year. I've worked full time as a Software Engineer for over 2 years now. I'm starting to think that being a code monkey just isn't what I want, but I love CS and I'm really interested in research. I did excellent as an undergrad (almost a 4.0), and received many merit-based awards, but it was a pretty small/unknown school. I know I can get some great LORs and I'm confident that with enough prep I can get at least a 1400 on the GRE.

My question is, do I have a reasonable chance at getting into a top school with no publications or real research experience? I have a few months now to prepare if I want to apply for Fall 2012, but my time is extremely limited (I work up to 70 hours a week) and I want to have some idea of what my chances are. Also, would 2+ years experience in the CS industry help my chances much?

Getting in a Big-4 schools would be a dream, but from what I've read, I wouldn't have a chance in hell w/o publications. Other than those, I am most interested in:

Princeton

Harvard

UNC Chapel Hill

Duke

Georgia Tech

Yale

UT-Texas

Johns Hopkins

Also, ML and Software Engineering are what I am most interested in studying.

Thanks in advance!

I think your lack of research experience will definitely put you at a disadvantage when applying to these top schools. Most of the candidates considered will have excellent GPAs and GRE scores, so you can't really depend on those too much. Your work experience may work in your favor when applying to programs focused on Software Engineering, though not as much as prior research experience would. Have you tried doing research? Are you sure that is the direction you want to go in?

[barber5]

Hi all,

I graduated with a double B.S. in CS and Math last year. I've worked full time as a Software Engineer for over 2 years now. I'm starting to think that being a code monkey just isn't what I want, but I love CS and I'm really interested in research. I did excellent as an undergrad (almost a 4.0), and received many merit-based awards, but it was a pretty small/unknown school. I know I can get some great LORs and I'm confident that with enough prep I can get at least a 1400 on the GRE.

My question is, do I have a reasonable chance at getting into a top school with no publications or real research experience? I have a few months now to prepare if I want to apply for Fall 2012, but my time is extremely limited (I work up to 70 hours a week) and I want to have some idea of what my chances are. Also, would 2+ years experience in the CS industry help my chances much?

Getting in a Big-4 schools would be a dream, but from what I've read, I wouldn't have a chance in hell w/o publications. Other than those, I am most interested in:

Princeton

Harvard

UNC Chapel Hill

Duke

Georgia Tech

Yale

UT-Texas

Johns Hopkins

Also, ML and Software Engineering are what I am most interested in studying.

Thanks in advance!

I'm sorry if you've already specified, but are you interested in getting a master's or a PhD?

One thing to consider: you are saying you know you can get some great LORs, but it's important to know that a letter saying "this person did great in my class" is not necessarily a great LOR for many top places including the ones you've listed and I can only assume that's what yours are going to say based on what you've said, correct if I'm wrong. Likewise with the GPA and GRE, while doing somewhat well here is necessary, it is far from sufficient. I only point this out because you say you're planning to do some prep to get a 1400 GRE--I tell anyone spending much time studying for it, your efforts would almost surely be better placed doing some research or improving your SOP. And like you said at 70 hours a week it's going to be hard to hugely change your resume before fall.

If you have suddenly discovered you have a passion for research and/or graduate work, I think you definitely sound like you are qualified to find a less demanding job that still pays well for the next year in order improve your chances for fall 2013.

[josh-cs]

I think your lack of research experience will definitely put you at a disadvantage when applying to these top schools. Most of the candidates considered will have excellent GPAs and GRE scores, so you can't really depend on those too much. Your work experience may work in your favor when applying to programs focused on Software Engineering, though not as much as prior research experience would. Have you tried doing research? Are you sure that is the direction you want to go in?

I did some "additional work/projects" as an undergrad that professors might consider research, but it was mainly me trying to independently figure out what others have done (especially in my math research class). The university was a teaching school, so undergrad research opportunities were few and few between.

Am I sure? I honestly can't say 100%, but that's because I am making a decent living with my B.S., but there's no real challenge, I don't think I'm doing anything that wasn't already done in CS a decade ago. I've read some literature about what life as a PhD student is like and how different it is from undergrad [1]. But I do know my passion is CS and I do have a lot of ambition and desire to work on tomorrows technology.

I do think that I'd rather go to a great school or not go back at all. I've lurked here for years, and many have mentioned how different the job opportunities are for the top schools vs the low- or un-ranked schools. My end goal is to work R&D in industry, not academia.

[1] http://www.cs.unc.edu/~azuma/hitch4.html (one of my favorite resources)

[josh-cs]

I'm sorry if you've already specified, but are you interested in getting a master's or a PhD?

One thing to consider: you are saying you know you can get some great LORs, but it's important to know that a letter saying "this person did great in my class" is not necessarily a great LOR for many top places including the ones you've listed and I can only assume that's what yours are going to say based on what you've said, correct if I'm wrong. Likewise with the GPA and GRE, while doing somewhat well here is necessary, it is far from sufficient. I only point this out because you say you're planning to do some prep to get a 1400 GRE--I tell anyone spending much time studying for it, your efforts would almost surely be better placed doing some research or improving your SOP. And like you said at 70 hours a week it's going to be hard to hugely change your resume before fall.

If you have suddenly discovered you have a passion for research and/or graduate work, I think you definitely sound like you are qualified to find a less demanding job that still pays well for the next year in order improve your chances for fall 2013.

I'm leaning toward a PhD. About my LoRs, I like to think my professors think more of me than just being an A student. I was always talking with them about the field, discussing CS topics that weren't even taught, talking about graduate school, etc. They know how dedicated, ambitious, and tenacious I am toward my work, so that's why I think I can get good LoRs.

I tell anyone spending much time studying for it, your efforts would almost surely be better placed doing some research or improving your SOP

Thanks for that. I've heard that before, but mostly my need to study for the GRE is to at least get that score that puts me in contention (Chapel Hill says their average candidate gets 90% in every GRE section, which is over a 1400 total, it's probably like this everywhere).

About getting some research experience, what's the best way to pursue that now? I don't have the guidance of a professor like undergrads do. Are there many paid research opportunities for graduates?

[newms]

You're right that getting into the top 4 schools would be extremely difficult without at least 1 publication, but the standard is not that exclusive for other top 20 schools - it is very possible to get into a top 20 CS school without a publication. You will have to demonstrate a focused research interest in your SoP and your letter writers will have to speak to your potential as a researcher and not just as a 'did well in class' type as barber5 says. Getting into a top 10 school is going to be hard without some research experience though. So the top 10 schools in your list - Princeton, Georgia Tech and UT-Texas will probably be harder than the others to get into. The others on your list should be more feasible, but you're going to have to narrow your research interest down from ML and Software Engineering. What aspect about ML interests you? What kinds of problems would you want to work on? Likewise with Software Engineering - you should focus your interest a bit more than this. I suggest that you read papers (or watch talks/lectures - academicearth.com and videolectures.net are good resources) in ML and Software Engineering so that you can focus your interests and be able to write knowledgeably about current research in your SoP. You also need to determine whether it's a MS or PhD you want - if you apply for a PhD without being sure of your intentions, it's quite likely that that uncertainty will come through in your SoP or other parts of your application, such as in the letters your recommenders write.

It's going to be hard to get research experience at this point, possibly you could seek out a prof a a local school that was interested in having someone assist them over the rest of the summer. I don't know if this would be feasible for you since you're working 70 hrs a week, but I've heard that it works for some people. FWIW I will be going to a top 20 school this Fall even though I don't yet have a publication and I had limited research experience as an undergrad.

Edited July 13, 2011 by newms

[csperson]

i went to a lowly ranked state school for undergrad and really wanted to go to a top school for CS PhD.

I had a ~4.0 undergrad GPA but no research experience. I was told it may be easy for me to get into a top 20 school for MS, but no chance for a top 20 school PhD. So the first time I applied to MS programs only and just received my MS at a top 20 school. I did a lot of research during my MS. Last fall I applied to a bunch of relatively good PhD programs and got into several of them.

So... You could just try to apply to a MS program at a good school, and gain some research experience. Then you may be able to get into good PhD programs easy. One disadvantage is that many top schools don't offer full financial support to MS students.

I have talked with several CS professors at top schools. Basically, they said if you have no research experience where you did something cool, then you are screwed for admission to the PhD program. They said they wanted to see recommendation letters saying you have research potential and describe the research you did kind of in detail. They don't really care about grades/GRE at all, unless you seriously screwed up. Lower ranked schools seem to really care about GRE/GPA though.

[barber5]

I'm leaning toward a PhD. About my LoRs, I like to think my professors think more of me than just being an A student. I was always talking with them about the field, discussing CS topics that weren't even taught, talking about graduate school, etc. They know how dedicated, ambitious, and tenacious I am toward my work, so that's why I think I can get good LoRs.

I tell anyone spending much time studying for it, your efforts would almost surely be better placed doing some research or improving your SOP

Thanks for that. I've heard that before, but mostly my need to study for the GRE is to at least get that score that puts me in contention (Chapel Hill says their average candidate gets 90% in every GRE section, which is over a 1400 total, it's probably like this everywhere).

About getting some research experience, what's the best way to pursue that now? I don't have the guidance of a professor like undergrads do. Are there many paid research opportunities for graduates?

Right, like newms was saying "research experience" for you might just be to read some papers and find something more specific you can get excited about/write about in your SOP

I don't know that it's generally advised to use MS as a stepping stone to PhD, but like csperson said you might consider it in your case. I know for a fact there are quite a few schools which don't formally fund MS students but where almost any MS student that wants it is funded (every MS student at UIUC that I know is funded, most Stanford students can find a TA after the first quarter I'm told, and I can't recall specifics but have heard of a few others), but I really don't know how to advise you on this because I haven't researched MS admissions too thoroughly--should you avoid terminal programs? will your industry experience help more for an MS app than PhD app? I'd think so, but maybe not.

Either way, if you're feeling trapped at this job, getting into an MS where many are funded may put you in a better position--if you don't get funded, you will probably be in a worse position. Maybe touch base with the professors who were fond of you and get some advice also.

[josh-cs]

Thanks everyone for the advice/opinions. I'm reluctant to try to use a MS as a stepping stone for a couple reasons. The main being the lack of funding for most places. And if I were lucky to get some funding, it'd be a TA and not a RA. Even if I got full funding with a decent stipend, it'd still be a big cut in income for my wife and me, but that's something we are willing to take. The other reason being that credits from institutions don't always transfer. Two of my professors got their MS from 1 program, then went to another for their PhDs and had to take most of the same classes over again. Many of my professors advised me that if I wanted my PhD, try to get straight into the program.

It seems like I'll have a slim chance of getting into even the #20+ schools (NC and Duke) that are actually near the top of my list due to reasons more than academics. Would it be strange/bad for me to contact some departments for more information about admissions, importance of research/publications, etc. that aren't posted on their websites?

newms getting into a top 20 school w/o a publication is a little encouraging. newms, do you mind sharing some of your credentials and anything else you think helped you get into a top school w/o publications and little research experience?

[newms]

Thanks everyone for the advice/opinions. I'm reluctant to try to use a MS as a stepping stone for a couple reasons. The main being the lack of funding for most places. And if I were lucky to get some funding, it'd be a TA and not a RA. Even if I got full funding with a decent stipend, it'd still be a big cut in income for my wife and me, but that's something we are willing to take. The other reason being that credits from institutions don't always transfer. Two of my professors got their MS from 1 program, then went to another for their PhDs and had to take most of the same classes over again. Many of my professors advised me that if I wanted my PhD, try to get straight into the program.

I agree with this. If you want to ultimately get a PhD, shoot for it now.

It seems like I'll have a slim chance of getting into even the #20+ schools (NC and Duke) that are actually near the top of my list due to reasons more than academics. Would it be strange/bad for me to contact some departments for more information about admissions, importance of research/publications, etc. that aren't posted on their websites?

No, it wouldn't be strange at all. In fact you could even contact profs, if you want, to determine if they'll have a need for new graduate students and to see if your interests match well with theirs. They may not respond to your email, but you never know unless you try.

newms getting into a top 20 school w/o a publication is a little encouraging. newms, do you mind sharing some of your credentials and anything else you think helped you get into a top school w/o publications and little research experience?

My stats were GRE: 800Q, 730V, 5.0 AW and my GPA would have been around 3.8x (I went to a non-US undergrad that didn't use GPA at the time, so that's an estimate on my part). I think what really helped me was that I focused my research interests in my SoP and my research interests were good fits at the schools I applied to (especially the one I am going to). What also helped me was that my likely advisor was looking to grow his lab, so he was looking to hire several grad students when I applied.

[curvature]

Hi joshcs,

I would say your situation is highly dependent on how well-regarded the professors writing your letters are.

It is difficult to over-state the value of a glowing letter from someone at the top of their field. I believe that their word can count for more than one's publication record. If you have people like this writing your letters, I think you will be able to get into the kind of PhD program you want with no publications.

Otherwise, I think you will need either: (1) a master's degree; or (2) a job that will put you in close proximity to a well-regarded professor and let them get to know you and your research potential. For the purposes of this discussion, I believe both count equally as vehicles to get glowing letters from well-regarded professors.

For example, here are two relevant job opportunities in top research labs, in case you are interested in computer vision. I suspect there are many more such positions that go unadvertised.

http://vision.stanford.edu/jobs.html

https://netfiles.uiu.../positions.html

For what it's worth, I currently work as a research assistant at an Ivy League school with a good computer science program. I work with a well-regarded professor in my field. I was not admitted to the PhD program at this school during the last admission cycle because I only started working there in the Fall, and my advisor didn't know me well enough when he was making admission decisions. The lesson here is that it takes time to build all professional relationships; your relationships with your advisors are no different.

The people that were admitted into my institution via my advisor fell into two categories. They either had: (1) stellar publication records and letters from very well-regarded professors; or (2) below average publication records and letters from very well-regarded professors. It is for this reason that I think letters from well-regarded professors can be more important than your publication record.

Indeed, I had a stronger publication record than some of the people in category (2). I believe they were accepted into the PhD program at my institution based on the strength of their letters. I doubt that their GRE scores made a difference. I base this opinion on informal discussions with my advisor, but you can check out the following blog post from Harvard computer science professor Michael Mitzenmacher for some more tangible evidence supporting this view.

http://mybiasedcoin....dex-impact.html

Finally, I think my comments will be more valuable if I disclose a bit more about my background. I have an undergrad and master's degree from a large (but not especially highly ranked) Canadian research university. My master's research produced a high-quality first-author publication. Unfortunately, my advisors during my master's degree are not especially famous, so I believe their letters (which I know were very strong, detailed, and research-oriented) were somewhat discounted when I was applying to top PhD programs. Anyway, with these credentials, I was able to gain admission to a top-50 American PhD program, and a top-5 European PhD program, but no higher than that.

Of course, your mileage may vary