General Acceptance Rates for Stat Phd programs

[Bayesian1701]

I am trying to get a ballpark figure on acceptance rates to middle and lower tier (30-60 US news ranking) Statistics Ph.D. programs.  I found data on acceptance rates on top 30 schools but not a lot on the middle tier programs.  I am a female domestic applicant with 3.75ish Math GPA (state school),  research experience, a possible undergraduate publication (if it gets accepted), and great rec letters.   I am guessing from survey results and the limited information I found that my chance of getting acceptance with funding is around 20% on average for these schools.  I want to apply to enough programs to get a 90%+ chance of at least one offer.  I figure you could build a binomial model of the number of funded acceptances. I want to know if 20% is a good estimate of the probability of me getting into a middle or lower tier program with funding. Obviously the probability varies across colleges, and this won't be the best estimate. 

 

Some Example Colleges : Florida State, South Carolina, Baylor, Virginia Tech, Kansas State

[Cioran]

You couldn't build a binomial model: the binomial requires both independence of outcomes and constant probability. This model has neither.

I think you're looking more for a logistic regression, which can map multiple categorical and quantitative variables into a binary response.

 

Otherwise, for this tier of program, the GRE score tends to hold the most weight in acceptance rates from what I've seen on the gradcafe and in talking with students from FSU and South Carolina. Nailing a Q: 165+ seems to all but guarantee you funding to those two in particular. Aim for that, and you should be golden.

Edited February 26, 2017 by Cioran

[Bayesian1701]

Given the same inputs getting into (or not getting into) school A should not affect getting into school B.  The probabilities are dependent on GPA, GRE, etc. but not on each other. The probability is not constant in real life but I am trying to get a rough guess. A lot of these schools have basically no public information on acceptance rates, and only a few survey results.  I'll look into the logistic regression.

[GoPackGo89]

If you crush this semester and end up publishing I don't see how you don't get at LEAST 2 funded offers from those 5 schools.

[Innominate]

Schools in this tier that I've spoken to have acceptance rates that are still around 5-10%, but some of them have higher international application percentages (70%+), and others just get a ton of applications that aren't anywhere near yours, so I wouldn't stress about the percentage.

You've left off two pieces of information which would help folks give you a subjective probability:  Your school and major/coursework.  State school isn't helpful, there's obviously a huge difference between UC Berkeley and New Mexico State, it does matter for your application odds (for better or worse).  You can remain anonymous by saying "unranked" or "top 50" or whatever.  Your coursework/major matters too, especially if you've had a ton of analysis or happen to have zero linear algebra, for instance.

Edited February 27, 2017 by Innominate

[MsHypatia]

Hi @BrittanyA1701. I wouldn't worry too much about probabilities or building a model. The problem is the acceptance rates schools report aren't always what they seem. For example, I got an email from a school today that said (paraphrasing), you are on our shortlist for admission, please reply back if you're still interested. Said school now doesn't have to report offering me admission (and risking being turned down) unless they're fairly certain I'll join their program. 

Do you see what I'm saying? The numbers can be fudged a bit. Not that I'm saying schools are dishonest. They're just trying to play the game to the best of the department's interest. 

I echo @Innominate. Post a few more details (ranking range of said state school, math courses and corresponding grades, etc.), and I'm sure at least a couple of people on this forum would be happy to help guide you in the right direction. Also, I would suggest looking at past "admissions results" threads and weighing your profile against others/comparing schools they were admitted to/rejected from.