Question about Safeties

[Edotdl]

For someone aiming for top 5 programs (ie MIT Stanford Berkeley Caltech for EE) what range of schools would be appropriate to be considered safties or matches? top 10? top 20? Thanks.

Edited October 19, 2015 by rising_star
to fix typo in title

[pterosaur]

There aren't really "safeties" in the undergrad sense when looking at PhD programs. In large part this is due to fit; even if a university is ranked lower, if your research interests don't match theirs you could very well be rejected. Fit is going to make more of a difference than the gap between top 5 and top 10. I honestly have no idea what the rankings are of the programs I'm applying to. My safety is going back to my undergrad institution, where my research advisor has said she would take me on for a PhD.

EDIT: Just looked up the rankings for the schools I'm applying to. None of them are top 10 for bioengineering. But then I looked up the top 10 schools, and none of them would be good choices for what I want to study. *shrug*

Edited October 21, 2015 by pterosaur

[EdNeuroGrl]

I second the above. 

There is no good safety school. I have a prof from my undergrad institution who would be willing to take me on, BUT the research fit is terrible. She would support me in my interests, but I'd have to do everything (including get funding for MRI level research) as well as some how forge a lot of collaborative relationships that wouldn't be there. These are things I CAN do but frankly, I want to spend my PhD work in an environment where I can be learning and being around people who care about what I do. I've already spent too much time having my research interests "tolerated" because I do high quality work.

 

So... programs fit or they do not, you have the opportunity to do your work or not. Your safest school is going to be one with a close match, with good funding, and bonus points for getting the PI to like you before you apply. That last one you don't have a bunch of control over so...

 

The best way to get rejected is to apply to schools based on rank alone. If there is no match and the PI's don't think you would be successful in their program, either because you don't have the same interests or it is clear you aren't considering anything but rank... you don't stand much chance. :-/

 

I am treating prestige as a tie breaker IF I get MULTIPLE offers and if all other factors are equivalent. So it doesn't  even enter into my consideration until much later.

[Edotdl]

I understand that fit is the most important, but that's not exactly what I mean when I refer to safties. I probably should have clarified, so my bad.

Anyways, correct me if I'm wrong, but I'd assume that given that the fit is reasonable (i.e. research matches your interests in both schools), a higher ranked school will be (relatively) more competitive than a lower ranked school.

So restated, assuming all schools have similar research fit, what range of schools would an applicant have high probability (70%?) of acceptance given they are competitive (3.7+ gpa, 330+ GRE, 2+ years research including summer programs (eg. REU's)) for top tier PhD programs.

Hopefully this will result in some different answers.

 

[EdNeuroGrl]

I'm not sure if there is a reasonable answer to this question. I've wondered myself, but if you wanted to get hard-core stats on it you would need some data about their past behavior as well as some objective measure of fittedness,which we don't have. We can't assume that this is a consistent variable among schools either since some very popular schools with get more applications with a much greater variability in fittedness ... *shrug*  There are a lot of factors that we just can't account for and we would need some base rates even for a Bayesian type of calculation...

So even given your assumption, as I see it your question about range of schools can't really be answered. Other than in a broad sense that it is smart to apply to several schools of differing ranks.

You could hypothetically apply to every program and have the best possible chance (assuming your application quality wasn't impaired), and maximize your chances. But, where is that probability versus effort at the most efficient? I have no idea how to even approach that question... If you talk to a bunch of people who have been through this successfully, you'll hear a lot of stories where they get rejected from 6-8 programs and only one accepts them and the ranking has nothing to do with it. Yes, higher ranked programs are harder to get into, but I'm not sure that the amount of variance due rank even approaches meaningful when compared to quirks of the program/POI funding/who happens to look at your application. 

There are many occurrences of people getting rejected from many low-rank schools only to get in at a top-rank school, AND many times where someone gets in to a low-ranked program and rejected from a bunch of tops, AND students with literally perfect stats who get rejected summarily, etc

Something to think on: often times rank will have something to do with how many applicants they get and how "high" they put the "minimum cut offs", but from there if you make it and you assume everyone has similar fit... there are still more applicants than spots (usually). It is also more common than you'd thing that PI's to get a list of applicants who pass the cut-offs, and those PI's have to nominate from the pool,  students that get reviewed by the department for consideration. So aside from making the "cut" you've just got to have luck or some other line-in on the PI to be considered. 

 

So, I'm not entirely sure that rank of the school is going to help inform your decision of which and how many schools to apply for.

 

I would like to know if others have differing opinions.