Advice - number of reach schools, safeties, etc.

[glasscandie]

What would you suggest be the balance between reach vs. safety vs. maybe schools?

I have 11 schools on my list (I imagine that will decrease, as I begin contacting researchers), 5 of which I've identified as reaches; 2 maybe's; and 4 safeties. I'm additionally applying to 1 masters program which I'm positive I can get into as a backup plan.

Do those sound like normal numbers? I'm afraid I have too many reaches (they are: (1) U Maryland College Park, (2) U North Carolina Chapel Hill, (3) Washington University in St. Louis (WUSTL), (4) Tufts, and (5) Boston U).

[aec]

I don't think there can be any hard and fast number or algorithm. In general I think it's a good idea to have the bulk be in programs you think you fit well with and have a good shot at, then 1-2 safeties if you have the money and 1-2 reaches if you have the money.

[jackassjim]

I went for many more reaches that safetys. I guess it depends how risk averse you are. But I figured that I have good stats, good letters and everything, and that my 10% chance at each of my 5 reaches gave me about 50% total...

Edit: Do I need to revise my GRE math, or is that ok?

[t_ruth]

I'm still baffled by the idea of 'safety' schools for Ph.D.s (applying to your current school/program aside). I can't think of a school that I would have a 100% chance of getting in where the degree would be a valuable one (as I want to go on and gain a professorship somewhere).

My distribution was one uber reach, two really top schools, four solid programs and two good but not as good programs. Don't think I'm guaranteed any.

[Joel418]

One other thing to think about when it comes to "safety" schools:

If you are applying to the school because you think the admission standards are more relaxed or that the competition for spots in less fierce, are you sure that it's a school that is going to be able to compete in your job market after graduation? I will certainly admit that there are programs to which I'm applying that have a reputation for more competitive admissions, but I wouldn't consider applying to a program whose reputation would keep me from getting past the "paper phase" of the job application process (the review of initial documentation such as resume, educational history, etc).

Personally, I'm a bigger fan of the numbers game, just applying to a good number of good programs, hoping that one will think I'm a good fit. I've applied to two Ivys, a school that models itself after the Ivys (Brandeis), the U of Chicago, and about five strong programs at state schools.

[frankdux]

I went for many more reaches that safetys. I guess it depends how risk averse you are. But I figured that I have good stats, good letters and everything, and that my 10% chance at each of my 5 reaches gave me about 50% total...

Edit: Do I need to revise my GRE math, or is that ok?

oh, if only that were so! then you could apply to 10 schools and guarantee yourself a spot somewhere!

Probability of getting accepted to at least one = 1 MINUS (prob of NO acceptances)

if all of the events have equal probability then we can say:

Probability of getting accepted to at least one = 1 MINUS (probability of NOT getting accepted on one attempt)^(# of attempts).

ex: probability of getting at least one Heads on 2 coin tosses = 1 - (0.5)^2 = 3/4

does it work? what are our possible outcomes: [H,H], [H,T], [T,H], [T,T]. and yes, 3 out of 4 of these have at least one H.

so in your case, lets say you are applying to 5 schools, each with a 10% likelihood of accepting you. therefore,

prob of getting accepted to at least one school = 1 - (0.90)^5 = 1 - .59049 = .40951.

so basically you have about a 41% chance of getting accepted to at least 1 school.

[IvyHope]

Thanks Frankdux! I've been wondering about that formula (NOT a math person).

[dragynally]

You have to do what feels right for you. You picked the schools for some reason and chances are its a good one. The stories of application to acceptance are all over the place. some people get rejected from safeties others apply to one school only and get in. You never know.

[glasscandie]

One other thing to think about when it comes to "safety" schools:

If you are applying to the school because you think the admission standards are more relaxed or that the competition for spots in less fierce, are you sure that it's a school that is going to be able to compete in your job market after graduation? I will certainly admit that there are programs to which I'm applying that have a reputation for more competitive admissions, but I wouldn't consider applying to a program whose reputation would keep me from getting past the "paper phase" of the job application process (the review of initial documentation such as resume, educational history, etc).

Personally, I'm a bigger fan of the numbers game, just applying to a good number of good programs, hoping that one will think I'm a good fit. I've applied to two Ivys, a school that models itself after the Ivys (Brandeis), the U of Chicago, and about five strong programs at state schools.

I actually didn't choose the schools based on safe vs. reach. I originally had a list of about 25 schools, and did a whole bunch of research on them, talked to some professors, etc. Then I cut the list down based on what I thought my chances were of getting in. For example, U of Maryland Baltimore County asks for a 3.0 GPA and their GRE scores are about average (high 500 area for V and Q); as opposed to say, Johns Hopkins where the average GPA is 3.7+ and GRE's are around 700's. All the schools that are on my list are schools I'd want to go to. I know it's really a crap shoot, I could get rejected from my safety school and accepted at one of my reaches. I always hear about people applying to only Ivy League or something and getting rejected from all.

I personally think I have a strong application - 2 internships at Johns Hopkins, 1 internship doing clinical/research, 2 independent theses which will be finished this year, 3.7 GPA, good LOR, and a publication. I've been talking to researchers at the schools I'm looking at and have had some great feedback. All the schools on my list I'd be happy to go to, but some are more competitive programs than others - which is what I mean by reach vs. safety. I didn't think that this was an unheard thing to do, am I wrong?

[jackassjim]

Frankdux: Thanks for posting this. Of course, I much prefer my formula. Nevertheless, I will defer to your expert treatment of the problem in this case. I will just need to revise (upwards) my estimated chances of getting in at individual schools so that I reach the same 50%.

[Ferrero]

Don't apply to a safety if you can't picture yourself going there and doing what you'd like. But it doesn't matter, really: CHE reports that people taking the GRE are way down this past year. By my lights, there will be plenty of spots at the top schools for even the most mediocre of applicants. Those applying to graduate school for next fall may in fact comprise only the active members of this message board.

[Tritonetelephone]

What I did was first find schools based on fit. If they had less than 2 faculty members that I would love to work with, they didn't make the list. I looked at their funding stats and made a few more cuts. Then I narrowed the list down to make sure that I had at least 2 reach schools, 2 medium schools, and 2 safety schools. So my safety schools are still a really good fit - which is really what matters in the end. FSU may accept 76% of applicants and 2 of my recommenders know people there, but I'm applying because I loved at least 6 of their faculty members (and they fund). I think that's the best way to get the maximum quality out of your graduate work.

ADD: Also, keep in mind that the "top 10" ranks for your subfield might not be in the "top 10" list for the field overall. That's how I found schools that would be good matches initially - FSU is #10 for sex/gender but #39 for sociology overall.

[leogk]

To glasscandie.

It's my tiny advice from last year experience.

Do not believe professors.

Positive feedback from them does mean nothing until you've got admission letter on your hands.

I had been heard a lot of positive comments in last year application, but the results were horrible.

(That may be partially caused by my major CS, which is one of most competitive fields to get in.)

The point is that you are certainly a very good candidate but there are five times more strong candidates like you than slots.

It's like a coin flipping...

[timuralp]

All the schools on my list I'd be happy to go to, but some are more competitive programs than others - which is what I mean by reach vs. safety. I didn't think that this was an unheard thing to do, am I wrong?

I'll only say that when I applied I got into a #5 school in the field and got rejected from #13 and #15, both of which had lower standards, etc. The problem is that when you apply, the department may not be looking for students in a particular area or may just not have the funding, albeit you could be a great candidate... It's hard to judge those things. In a lot of schools admissions are a crapshoot, but don't let it deter you.

[glasscandie]

To glasscandie.

It's my tiny advice from last year experience.

Do not believe professors.

Positive feedback from them does mean nothing until you've got admission letter on your hands.

I had been heard a lot of positive comments in last year application, but the results were horrible.

(That may be partially caused by my major CS, which is one of most competitive fields to get in.)

The point is that you are certainly a very good candidate but there are five times more strong candidates like you than slots.

It's like a coin flipping...

I'm not counting on anything!

[glasscandie]

I need to add, this is really flipping hard, cutting down schools! My research interests are pretty specific (social communication/cognition in animal subjects via the nervous system - which is generally neuroethology; and I'm additionally interested in how environmental factors i.e. iron or antioxidants would weigh in). The most basic way I've been doing a search for schools is googling different key words with "edu" in them. Haven't had too much success elsewhere.

[jackassjim]

The most basic way I've been doing a search for schools is googling different key words with "edu" in them. Haven't had too much success elsewhere.

Really, the most basic way should have been to speak with your profs...

[glasscandie]

That I have also done. I didn't say it was the only way I researched schools.

[rising_star]

Really, the most basic way should have been to speak with your profs...

Not everyone has professors to ask...

[grex]

oh, if only that were so! then you could apply to 10 schools and guarantee yourself a spot somewhere!

Probability of getting accepted to at least one = 1 MINUS (prob of NO acceptances)

if all of the events have equal probability then we can say:

Probability of getting accepted to at least one = 1 MINUS (probability of NOT getting accepted on one attempt)^(# of attempts).

ex: probability of getting at least one Heads on 2 coin tosses = 1 - (0.5)^2 = 3/4

does it work? what are our possible outcomes: [H,H], [H,T], [T,H], [T,T]. and yes, 3 out of 4 of these have at least one H.

so in your case, lets say you are applying to 5 schools, each with a 10% likelihood of accepting you. therefore,

prob of getting accepted to at least one school = 1 - (0.90)^5 = 1 - .59049 = .40951.

so basically you have about a 41% chance of getting accepted to at least 1 school.

That's assuming all the events are independent.

If, for example, the same random people were applying to the same 10 schools, and all 10 schools had exactly the same admissions criteria, you might have a 10% chance getting into any of those 10 schools depending on your chances of being better than 90% of the applicants. Getting rejected from one school, though, would mean you would have 100% percent chance of getting rejected from the 9 others.

Of course, the above scenario is an extreme. In general, admissions are not as correlated, but still pretty correlated. So your probability of getting accepted into at least 1 school, I'd imagine, would be under 41%, maybe significantly so (assuming your 10% estimate was correct).

[frankdux]

that's an excellent point.

however there are plenty of stories, even right here on these boards, of people getting accepted by one of their reach schools and being rejected by their safeties. the correlation that may exist is probably somewhere in the middle.

[jackassjim]

grex,

To elaborate a bit on your point...

What we are really talking about here, is not so much a correlation between variables, as the stability (standard deviation?) of the ranks that different admission committees assign to a particular applicant. A low SD (high correlation for you), by definition, means that there is little spread around the mean rank that is assigned to her.

A shorter spread is very bad for those who consistently rank in, say, the bottom 25th percentile, because it means that they are very unlikely to benefit from an 'outlier' ranking that goes in the right direction (down (up?) toward that #1 spot). The 'correlation effect' you discuss is not nearly as problematic for those who are very competitive, i.e. those that have a mean rank that is very close to the acceptance threshold. In effect, it is unlikely that the association between rankings is strong enough to shrink the spread so that those applicants are kept out of every schools. Of course, the converse of the argument you make in your post is that, if you are generally above said threshold, you want much higher correlation/lower SD. In other words, it's all relative to where you are in the pile.

Now, the question of red flags is something else. If there's one of those, you could very well be rejected everywhere. I guess the 10% already assumes there is no such deal breaker.

Edit: I see how my 10% estimate might have induced you into believing that that hypothetical candidate was near the bottom of the pile, in which case your intuition is perfectly fine.