Perspective from other side of the admission committee

I am a professor on the admissions committee at a medium sized T20 engineering department in the US and wanted to share some honest perspectives from the other side, as we often aren’t allowed to explicitly answer certain types of applicant questions. For example, many applicants want to know our acceptance rate which are not supposed to share. My program accepts roughly 35-40 students out of 600+ applications, and our yield tends to be somewhere between 50-60% of those admits join the program.

Our process: the admissions committee reviews applicants and ranks them on a score from 1 to 3, where 1 is excellent, 2 is good, and 3 is unsuitable. Most applicants are fairly realistic about their chances of getting in, I would estimate roughly 10% get rank 1, 85% rank 2, and only 5% of odd cases are ranked 3. After that, the scores and application materials are shared with the rest of the department. We are a direct-match program (i.e., students get accepted directly to individual lab groups, rather than as a cohort), so individual PIs then get to decide who they will interview. The admissions committee will make notes of which professors should look closely at which applicants. Not every professor will have funding for new PhD students every year, so many applications (even excellent ones) are never strongly considered. Rank 2 applicants are sometimes accepted if the research fit with the professor is very good.

You may have heard this before, but there is no such thing as a safety school for graduate applications. We routinely reject rank 1 applicants simply because there isn’t a professor in their field of interest who has an open position that year. So having the best profile does not mean you will get accepted, you also need to get lucky that the right position in the right group is funded for you that year. For smaller, less research active schools, this means that there are often fewer positions available, so some of those programs may actually be harder to get into compared to larger and higher ranked programs like MIT, Michigan, and Georgia Tech which need to hire large numbers of students to support their massive research programs.

GPA matters. While research proficiency is most important for a PhD, a poor undergraduate GPA doesn’t bode well for your chances of successfully completing the pre-requisite coursework in a graduate program. These classes are hard, and if you are spending all of your time studying just to do okay, you won’t have time to start research and your chances of passing the qualifying exam will be lower. Many professors consider ~3.7 or above to be acceptable, but top applicants usually have 3.8 or above. I don’t say this to discourage you if your GPA is lower, but I also don’t want to sugar coat what type of profile tends to be accepted.

A question I see all of the time is: does research experience offset a mediocre GPA? The diplomatic answer you’ll get from most admissions staff is that applications are reviewed holistically and there is no minimum GPA. But the honest answer is: probably not. Several applicants will have both research experience and an excellent GPA, and in many cases the “superstar” rank 1 candidates will have a higher GPA in addition to more research experience than a rank 2 applicant with a decent GPA and some research experience. Out of the 100s of applications I have read, I can only think of one case where a candidate had a 3.2 GPA but such excellent research experience and letters of recommendation that the application was still strongly considered.

Another common misconception is the importance of publishing as an undergraduate or masters student. Having a publication can certainly boost your application, but it is far from a prerequisite. We routinely accept students who have no publications. Doing science takes time, and doing good science is usually especially slow. In fact, having your name on subpar publications might actually work against you. I was recently contacted by an international masters student who has more publications than me, because their father is a professor who has been adding their name to all of his (not very good) publications for the last 6 years. I am fairly confident that this super-obvious “gaming” of the academic system will result in this student getting rejected from all top programs. Then they will go to grad cafe or reddit and complain about how impossible it is to get accepted into graduate school if they got rejected despite having X number of papers. So don’t get discouraged if you haven’t published when you read those types of posts!

Another common question seems to be whether international students are at a disadvantage. The sad answer is yes. This is for a few reasons: (1) there are many funding mechanisms only open to US students (the big one being NSF GRFP, but there are several others), making it easier for professors without enough funding to accept them, (2) we know exactly what a 3.9/4.0 from the University of Delaware means, it might be harder to evaluate a 9.0/10 from IISc, (3) we are more likely to have a connection to, or know of, the professors at American universities writing letters for those students. The deck is especially stacked against Iranian applicants. Although there are many wonderful junior scientists in Iran we would love to bring over, the reality of visa delays/rejections and extra scrutiny means many programs/professors can’t or won’t gamble on making offers to those students. If you are international, don’t give up hope though! There aren’t enough excellent American students to fill all the US programs, so most top schools still end up with a majority of international students. You just might need to apply more broadly than an American student would.

Make sure to get your applications in on time, including letters of recommendation and IETLS/TOEFL scores. While exceptions might be made for superstar candidates, last year we weren’t even forwarded the applications that weren’t completed at the deadline. I had a few students reach out to me to ask if I’d seen their application, and I hadn’t because their IETLS scores were delayed and the admissions staff had only sent us complete applications.

My final thought is to make sure your personal statement reads well, especially the first few paragraphs. This is the first part of the application we look at and we generally make a judgement fairly early in reading. I try to do the courtesy of reading each statement in its entirety because I feel that we owe that to applicants who put so much time into applying, but the reality is that many professors will skim the statements and make a snap judgement since we are analyzing so many. If you aren’t a strong writer, use AI to help! AI writing tools can help level the playing field for non-native English speakers. However, do not copy and paste directly from chatgpt. It is incredibly obvious when someone has done so. Make sure the statement still has your distinct voice and thoughts and does not include generic wording that doesn’t tell us anything about you. Sentences such as “I love XX field because I have always liked math and physics” are true of every engineering applicant. I want to know more about you as a person, and every word you choose to include in this statement should help make your case. I realize that this is easy advice to give, and not easy advice to incorporate, but do your best to think about what makes you unique and interesting. Also, don’t be afraid to brag a about your accomplishments. If you have published, won awards, conducted outreach, etc., include that in your statement. Give us context for awards we may not have heard of (selected out of XX applicants), include metrics of impact (my outreach project was shared with XX number of low income students). Give us context to your research experiences (how long were you with a group, did you work alone or under a postdoc/phd student, what tools did you use, what were your main contributions to any resulting publications, etc.). And of course, have someone proofread. Sentences that make sense to you might sound like gibberish to someone else, which is why we often cannot effectively evaluate our own writing.

You can absolutely be accepted to a top program with a 3.2, and you can also be rejected with a 4.0. The last thing I want to do is discourage anyone from pursuing their dream program, but I also want to be honest about what types of candidates are typically accepted to top programs. For example, my last few years of admits:

• 3.5 UG, 3.9 M.S. International, 2 research experiences, 1 publication, 1 presentation, leadership experience, letter of recommendation from a professor I know and trust. SOP indicated very strong interest in my specific research field and as well as the application I care about
• 3.85 UG, 3.95 MS. International, 2 research experiences, 2 presentations and 1 in-progress publications (but not published), leadership and volunteer experience. Referred by trusted colleague, excellent research fit.
• 16.5/20 UG, 3.7 MS. International, 3 research experiences, 2 publications, significant outreach experience, amazing letters of recommendation from unknown professors. SOP indicated very strong interest in my specific research and application I care about
• 3.98 UG GPA. American, URM, 2 research experiences, no publications, significant outreach experience. Letter from a trusted colleague. SOP indicated very strong interest in my specific field and and application
• 3.8 UG GPA, dual major. American, URM, 2 research experiences, presentation but no publications, excellent leadership experience, referred by a trusted colleague. SOP a bit vague but good enough alignment with my research
• 3.9 UG GPA, american. Top UG program. Awards, 1 research experience, one publication, 2 presentations, volunteer, leadership, outreach experience, excellent letters from unknown professors. SOP reflected good alignment with my research, but not with my application.

You might notice a common theme is that referrals/letters from other professors I know personally hold a lot of weight. I have used the phrase “take a gamble” a lot in my comments, because that is what we are doing when we accept students. In between tuition, stipend, fringe, overhead, and research/travel costs, it costs over $100,000/year to train a PhD student at my institution. This is money we professors need to painstakingly fundraise. Because PhD positions are some weird combination of a job and a training program, making a bad hire can have an enormous impact on our research programs. It’s not like a normal job where I can just fire someone if they aren’t working out 2 months in. The last thing any professor wants to do is spend 200-300k training someone who ultimately isn’t productive and burns out early because they actually don’t care about the research area. This is why programs are so weird about “why us?” We want you to convince us that you will be happy and successful in this program and aren’t going to drop out. You might be the strongest applicant in the pile in terms of raw metrics, but if we don’t see the clear alignment of interests you may not be accepted.

Personally, I am also very interested in personality match. I don’t want to spend 5 years butting heads with someone because we have different priorities and working styles, and I especially don’t want someone who will make the rest of my group miserable by being a pain to work with. This is why I put a lot of weight into personal recommendations from people I know. By the time I am interviewing candidates, it’s really more of a “vibe check” than trying to assess competency. All professors are different though, some will really grill candidates for technical competency, which I personally find unproductive.

Finally, if your profile is not as strong as the ones I have mentioned, please do not despair or give up hope on doing a PhD. I am describing the admissions process at a very competitive top program located in a highly desirable city. There are many R1s with high research activities and plenty of funding that don’t make it onto top 20 lists. For example, state schools in “rural” states have access to a separate pot of NSF funding that coastal states do not have. The university of texas system has their own sizeable endowment. There are many excellent, T100 programs physically adjacent to top schools that are sometimes overlooked by applicants (i.e., NJIT near Princeton). Top schools located in less desirable locations will also be less selective. Because of the political climate there, colleagues from red / southern states have been complaining recently about not getting enough female and out-of-state applicants in their pools. Canadian programs have a very different funding mechanism than the US which results in more equitable distribution of funding across their various schools. Finally, try to find out if a program of interest has hired a lot of new professors recently, which suggests that the school has funding and potentially more openings for PhD students.

If you do decide to apply to top programs, make sure the alignment is clear in your SOP, and try not to take it as a personal failing if you end up not being selected. We all want to believe in a meritocracy with a fair and systematic process, but the reality is that professors making these decisions are just people and the system we use is sometimes arbitrary or downright stupid. We make mistakes, we overlook good candidates for stupid reasons or because of personal biases, we spend less time on applicants describing research interests far from our own fields, we forget to read the last 2 applications on the pile of 100s, etc. I know candidates often want to know “what was wrong with my application that I didn’t get selected?” but this is the wrong way to think about it because there may have been nothing wrong with your application. In reality, it was just that something in someone else’s profile that made them stand out to that particularly professor, such as a letter of recommendation from the right person.

I hope this helps, best of luck with your applications everyone!

*Copied from a reddit post for guidance.