I'm the 1st author of a paper in a top conference in my field, and even then my PhD application got rejected from 5 out of 6 programs, is this normal?

Question

It got me so disappointed because I thought the publication experience was the most crucial part of the application, and, as far as I could tell, not that many students in the universities I applied to got 1st author papers in top conferences before starting their PhDs. The conference where I got my paper had an acceptance rate of less than 17 %, with more than 8k submitted papers, and also among the top 10 in my field for H5-index and top 5 for impact score. So I wonder if this sort of thing is normal, to get rejected with a strong publication.

My field is Computer Science, focusing on Artificial Intelligence, which is in notoriously high demand these days. Still, I didn't expect it to be so competitive that I wouldn't get at least another accept offer than the single one I got, which was from my least preferred option.

I applied in the US, and 3 of the universities that rejected me are in the top5-top10 range in the USNews ranking for grad programs in AI (https://www.usnews.com/best-graduate-schools/top-science-schools/artificial-intelligence-rankings), and also the other 2 that rejected me are in the top10-top25 range in this same ranking. The one university that accepted me is not in this ranking.

To be fair, other aspects of my application were a bit weak, I must admit, for instance, low GPA for undergrad (even though I got excellent GPA for my masters), didn't send GRE scores (it was optional for the last application cycle), and my previous universities are outside of the US and somewhat unknown internationally.

Answer 1

What you find is that admission isn't based on one criterion alone, but it is a holistic process of looking at an applicant's entire file. This includes not just this one paper, but also your grades, prior research experience, the letters of recommendation, which university you have previously been at and how that university is perceived by others, and many other factors. You mention some of these aspects in your last paragraph, and those who have read your application must have considered these relevant in making their decision.

I think that that is also the right approach for making decisions. Your GPA is a cumulative achievement showing how you work over several years. A single paper, with a number of co-authors, is a noteworthy achievement but it only provides a relatively short-term perspective on your ability to work. It is of course a positive aspect of your application, but apparently was not enough to overrule other aspects.

Answer 2

Congratulations on being accepted. No, seriously. If you are about to start a PhD, and you have ambitions to go into academia, you need to come to terms with the fact that a large part of your professional life will now consist of sending applications for positions, grants, equipment, travel, tenure, promotions etc. If you are like most of us, the majority of your applications will be rejected. Most schemes in academia are highly competitive, with acceptance rates below 10%. Getting 1/6 with a less-than-stellar application means that the rest is good. Take the offer, and make the best of it.

To answer your question more directly: While a publication record is important in academia, it tends to be less important before you start your PhD. Whenever I've been on PhD hiring commitees, much more emphasis has been put on grades and recommendation letters than a possible publication record. Thing is: It is not really expected in most places that you have a publication record before starting your PhD, so if you have one, it is mostly just icing on the cake -- if there is no cake to begin with, you will not be hired even if you have several good publications.

And on a more personal note: I have on several occasions seen candidates with poor grades, and a handful of papers. In most cases they have given the impression of a weak candidate when interviewed. I should give the position to a candidate who will make a good PhD student, ie. take the courses and do well, and then do research. If you fail to convince me that you have somehow turned around, and now will start doing well in courses, I will not accept your application, regardless of how many publications you have. Also remember this: I cannot know how much you really contributed to that publication. It could just be written by a former supervisor, who I am not hiring.

Answer 3

It's been mentioned that the process is very competitive and a publication alone doesn't imply a strong overall application.

I would just add that PhD admissions are often about match to potential advisor research. It's not about whose application is most impressive, it's about who they want to do research with.

Your application may not have been read by the professors at those schools who are close to your interests. (Maybe you didn't signal well who you would like to work with.) Or you said you're interested in topic A, but your publication is more about topic B and professors who do A were not excited about your app. Etc.

Even more likely, the professors who share your interests might just not be hiring this year, or they may not exist at those institutions if your interests are niche.

Answer 4

I applied for MSc in Computer Science this year to 4 universities in Canada and got rejected from 3 of them (no decision yet from the fourth one). My GPA is 17.84 / 20, my last two years' GPA is 19.01 / 20 (sometimes the last two years' GPA is more important for universities in Canada), and I ranked 1st among 73 undergrad students. My IELTS score is 8.0 (less than 2% score 8.0 or higher in academic IELTS), and I have several projects and research in AI, image processing, and NLP (No publication, though).

There are a few problems; I guess because of covid in the past two years, many students decided to wait, and now that everything is going back to normal, they are all applying at the same time. Because one of the universities I applied to said they make all decisions by the end of April, mine didn't come, I emailed them, and they responded that since the number of applications was so much more than they expected (more than 1400) it took them a long time to make all the admission decisions.

Not to mention that the competition for computer science is unbelievable. Another problem is that, at least in Canada, no matter how much the faculty members say we don't make decisions and the admission committee will admit/reject you, they are the ones who make the final decisions. What is the problem with that? Try to contact them, send an email, and explain how much of a perfect match you are for their project/work/lab/research you get no response because when 1400 people pay more than 100$ and fill out a long form to apply, so much more will send an email, and your email will get lost.

When the number of applications is really high, they may automatically ignore a huge proportion of them. For example, the only weakness I see in my application is that I graduated from a 1200+ ranked university. Since they have so many applications, they might not even look at those who graduated from any university with a rank higher than 800. So no matter how much talented you are and how good your resume is, you don't even get a chance. Because let's be realistic, they are not going to review 1400 applications one by one, and this is just the MSc in Computer Science. Think about Ph.D. applications, bachelor's, and other programs in that faculty.

Answer 5

I'm a professor in ML/AI at a Canadian school that's probably roughly in the same caliber for AI as the #15-25 schools on the US News list. (I got my PhD at the #1 school on that list, but at a time when admissions in this area was far easier – I am quite confident I would not get in today with the application I sent in 2011. Even then, I was also rejected from some schools ranked far lower on that list than the one I went to.)

I don't have the numbers in front of me, but admissions rates in AI/ML are actually far worse than for CS grad programs in general: in my department, more than half of the applicants currently say they want to do broadly-AI topics, while even judging very generously as to who counts, under 1/5 of the professors are in that area.

It sounds like you have a first-author AAAI paper – congrats! This is definitely a positive signal that makes me pay more attention to an application. Depending on what area of AI you're talking about, though: "core ML" people like me tend to value AAAI papers significantly less than those from venues like NeurIPS, ICML, or maybe ICLR/AISTATS/UAI, all of which have higher numerical acceptance rates but whose average ML paper tends to be (in my opinion) of higher quality than the average AAAI ML paper. (This is not the case for other areas of AI, though, where AAAI is the premiere venue.)

But: you say you've already done a master's. Having a paper out of undergrad is definitely a distinction from many applicants who do get in (though not everyone agrees), but out of a master's, it's far more typical. (I would expect that most, though certainly not all, admits to top graduate programs who already have a master's will also already have a paper or two.) It's for sure a good thing, but not something that sets you super apart from the competition.

So, this leaves all the other factors that everyone is talking about. The middling GPA from not-well-known institutions certainly hurts. Maybe your reference letters weren't as positive as they could be, particularly if they're from people and/or a country that US schools don't know so well. Maybe the topic of the paper and/or what you wrote about in your statement is different enough from potential supervisors' interests that they have a hard time seeing themselves working with you.

Candidates who are "pretty good but not super-outstanding" can tend to have a hard time in admissions, since the strongest-on-paper applicants will tend to get a lot of offers, but advisors are wary of over-offering since then if too many students come they'll have to find a way to pay for them all.

(As you've mentioned elsewhere, it's highly unlikely you're going to get any specific feedback on that from anywhere.)

In any case: it is what it is. You can either go to the school you're not so thrilled about, or decide to do something else with the few years / rest of your life. The quality of the work you do in grad school is more important than where you do it, and though this will make things more challenging versus going to a top place, this is definitely not necessarily the end of your academic career or anything – plenty of excellent researchers went to lesser-known schools.

Answer 6

Rather than questioning your application and suitability, have you considered who your competitors are? In my experience of applying for and interviewing for a PhD I found that the position would frequently be offered to people who already knew and had worked with the supervisor(s). Advertising the PhD publicly is a funding requirement in a lot of cases. When I eventually did get accepted to a programme it was a result of contacting my future supervisor directly, then seeking funding. This is the case for many PhD students I have met. So, I do think it is normal to be rejected so often.

Answer 7

It's very normal. Even if you had a stronger undergrad GPA and good GRE scores it would probably still be normal.

Congratulations on being accepted. A single paper isn't going to do much for the acceptance committee. If you'd like to go to a more prestigious school, then you should take the GRE, do very well, and reapply.

Answer 8

You should thank the admission committee for not admitting you to their PhD program. This is the sanest thing to do when no professor is interested in further supporting your advancement in the program after you get in.

I was admitted to a top 50 CS PhD program in the US without a pre-assigned advisor. After one year of futile search for a PhD advisor in the program, I could not find one and had to leave the program at the end of the first year, as department discontinued my funding. I was an international student coming from a poor country and it was impossible for me to afford the tuition and other costs.

During my study in the program, almost all faculty (around 14-15 people) relevant to my intended research field (very broadly AI/ML) did not even consider responding to my emails that inquires for positions in their labs. Note that these emails were sent after I gained admission and was officially a student in the program. I could only get polite rejections after graduate coordinator got involved in the process and asked faculty to reply to my emails. I did not lie or misrepresent my research interests, professors whom I am interested in working with etc. in my graduate school application. So at the first place, I should not have been admitted to that PhD program. From the day one, it was obvious that no relevant faculty was interested in my presence in that program.

In addition, my experiences with other students indicated that lots of communication happened between professors and students long before the admission process. Admittedly, some students were much better than I was in terms of research experience and publications; but some of them had no prior publications or a serious research experience, but had the same nationality with the professor. Probably, school factor, coming from the same culture and alma matter issues heavily influenced decisions of professors. Even during discussions with senior PhD students, some suggested not to even consider applying to (non-American born) Chinese professors' labs because they never admit non-Chinese students. Whether you consider it as a form of discrimination or having a better ability to evaluate students from your own culture, usually professor’s care more about sharing cultural values, nationality etc. Consequently, recommendations from their colleagues in their home country (India, China etc.) might overwrite unknown someone with a good publication record.

I suspect the program heavily needed a large number of cheap teaching assistants to compensate increasing undergraduate enrollment and my PhD admission was a disguise to cover up that demand and to exploit me as a teaching assistant, rather than seeing me as a research student. Otherwise, you do not admit a student that no professor wants to supervise. My firsthand experiences indicate that US PhD programs are no way fully meritocratic, but this is the image they present to the world. That's why you think that your publications should get you in.

I wish professors could openly say "do not apply if you are not from country X, school Y or race Z". This would allow both ends not to waste their times, but discrimination laws in US do not allow that so they had to revolve around it in other ways. It has to be projected as your inadequacy.

Answer 9

I recently helped screen AI/ML applications at one of the top-4. Unfortunately, having a publication is essentially a requirement, not a plus (this isn't necessarily true for non-AI or other universities). We were also told to expect more from masters students than undergraduates. For instance, my job was to narrow it to the top 10% (to be narrowed further by PIs) and I remember feeling very bad about having to reject a masters student with 2 good publications and good application package overall. PIs knowing the letter recommenders also helps a lot. It's indeed harder for countries that do not encourage research from early on, as the US does and from lesser-known universities.

Having advised many undergrads while in grad school and talking with my PIs about this, some of the main criteria for choosing whether to go to grad school should be:

• Do you want to be a grad student? If this answer depends on going to the top-4 (and not at least top-30), the answer is probably no.
• Do you like AI so much more than other topics? The answer may be no, in which case going to related areas may give you a higher chance at cracking the top-4. Still, if you do not want to go to a non-top grad school I wouldn't do it.
• Do you like an advisor at that university? There are plenty of great advisors outside the very top universities. He/she will be your main academic influence by far. Being at a top university just provides a larger safety net of alternative advisors, as well as more people to get advice/conversations from time to time.

Answer 10

Consider treating your Ph.D. application as a job search process - which it is; you would be working as a junior researcher. You've just described to us how you applied to a bunch of universities as generic entities, with no prior contact with the potential employers. It's a bit like sending your CV to arbitrary companies, asking "Would you like to hire me for a junior position?" - Instead of going at it this way, you could have thought about potential subfields you are interested in, found some relevant academics working on that at a decent university, got in touch with them, and if it seems there might be mutual interest - then apply, or rather have them promote an application for your admission/employment.

_{(This is not to detract from the other valid and useful answers.)}