Reviewing grad applications

[EngineerGrad]

Hi guys,

 

Today my advisor gave me a stack of applicants for our lab and asked me to select my top 3 (we will hire one phd student to work with me in collecting data for my/our research). Have you guys done this before? I don't know how to choose them...

 

There's one applicant that I really like but his SOP is soooooooooo bad. Since I'll need someone who writes well, I don't know what to think about this applicant.

 

Any advice?

[rising_star]

I've never been asked officially to do this. But, I would think about things like research experience, GPA, SOP, research fit, and personality, especially if I'd be working closely with this person in a lab for several years. You want someone that's smart, has some research experience or is eager to learn, and has interests that align with the lab and the work you're already doing or plan to do. I have GPA in there but, having a high GPA doesn't mean someone will be good at research just like a low GPA doesn't mean someone will be bad at research. Hope this helps!

[fuzzylogician]

Here is how I imagine doing it, though disclaimer: I've never had to actually do anything like this.

- First pass: look at research fit/interests, training. Can this person do the job? This might involve scanning the SOP, CV, and possibly LORs but only for a description of interests and relevant background. 

- Of the people who have fitting research interests, do they have the skills? This might involve looking through transcripts, SOP, LORs. If writing is important, at this point I might pay more attention to the SOP and writing sample, if you have one. If they are lacking some skills, does it look like they are able to learn new skills and apply them quickly? Are they eager to do that and express and interest? 

- Of the people that have the right skills, do they seem like hard workers? This might involve looking at GPA and GRE scores, though I wouldn't trust them too much; also looking at what the person has done - involvement in projects, any presentations that have come out of previous work, an engaging SOP that spells out excitement and an genuine interest in what I do. 

- Finally, the intangibles. Do they seem like a good personality fit? This may be hard to know but based on SOP and LORs there might be some indication. It'd be helpful to be able to talk to top candidates on the phone, though I realize it may not be possible. Hopefully by this point there aren't that many applications left standing, though, so you can pay attention to the content of what they did submit. This might also be the time to google them and see what comes up. 

 

Writing this out, it seems potentially time consuming, but personally I think it's worth it for someone I'm going to work with for several years and that my research will depend on. Then again, who knows, I'm sure there are more efficient ways to go about this.

[Eigen]

Cross posting this here, since it's got good parallel advice.

http://chronicle.com/forums/index.php/topic,176933.0.html

In general, I'd consider writing ability important, as well as past research. Personally, I lean towards someone win a variety of past experience, and a little more maturity, but that's hard to tell.

Low GPA can be a flag, but a high GPA doesn't necessarily mean anything, IMO. Research experience and evidence of work ethic, however, are harder to see.

I'd probably base a lot on letters, especially ones that have a good personal tie to the student.

Are you going to be able to do interviews? I'd pick a couple that look good on paper, and see if you can do a phone/skype interview.

So many experiences with people who look great on paper, but I wouldn't want to train or work with.

[SNPCracklePop]

I would lean towards those with past research/technician experience, the more the better.  In your case, from what I understand, I would not too concerned with the SOP.  Scientific writing is a skill that can rapidly be improved, especially if the applicant has a firm grasp of the language.  If the applicant is being hired to assist you with data collection, it sounds like writing will be in the minority of their duties.  If that's the case, a little proofreading and editing by you or your advisor shouldn't be too taxing.

 

The LORs will be useful here, and you should certainly contact references if you need more information.  They can give you insight on their personality, work ethic, etc.

 

I would be tempted to do a quick Google search if you are still undecided after narrowing down the candidates.  Good luck!

[thr0waway]

if i were in this situation i would google them / check out their social media profiles before anything else

 

as a coworker you don't need to care *that* much if they're a good student.  you should make sure they're reasonably competent (so they don't ruin your project), but they don't need to be a superstar who'll publish 10 papers before graduation

 

their main impact on you will be social, and if you're going to spend a lot of time in lab with them, you'd better make sure they have a good personality.  let the professor do the academic filtering; he's the one who benefits most from having good students.

Edited February 6, 2015 by thr0waway

[TakeruK]

I agree with thr0waway in that it's not up to you to choose the best student for the program etc. That's your professor's / department's job. Pick the student that you think will be the best co-worker for you. Make sure you would get along well with them and make sure that they have the specific skills you need to best assist you in the project. I would not worry about things like "will they be able to pass the courses?" etc. because that's the department's job.

[juilletmercredi]

Disagree.  You've got to think bigger picture here.  Of course personality and work skills are going to be the first considerations, but of course it matters to you whether or not the student is a good one and/or is the type to publish 10 publications before graduation.  First of all, on the most concrete level, a student who comes ready to publish can work with you to publish papers - i.e., you can work together and be first author on some and second author on others. Good students also foster better discussions both in class and in the lab, which may stimulate your own ideas and better work out of you.  At more abstract levels, more publications makes your lab look better and your department look better, and a superstar lab mate becomes a well-placed networking contact in the future.

 

The other thing is that you have to look at this task from the perspective of the person asking you to help with it.  Your advisor asked you to help most likely because he trusted your judgment; if you keep a narrow view towards this and pick students solely based upon their technical skills (which I know wasn't the advice but stay with me), your advisor might be...skeptical about your judgment.

 

You aren't hiring a lab technician or an undergrad RA; your advisor is asking you to help him bring on a graduate student.  So you should be concerned with selecting a grad student - yes, someone who can help you in the lab, but on a broader scale someone you would want as a colleague both right now in your department and in the future as part of the field.  That's what faculty members consider when they bring on people and you should, too.  Think about if you wanted a grad student - who would you choose?

 

Besides, you can't tell personality from pieces of paper or even social media sometimes.  I rarely post on social media and someone who checked mine wouldn't be able to find out anything meaningful about my personality from them.

[TakeruK]

You aren't hiring a lab technician or an undergrad RA; your advisor is asking you to help him bring on a graduate student.  So you should be concerned with selecting a grad student - yes, someone who can help you in the lab, but on a broader scale someone you would want as a colleague both right now in your department and in the future as part of the field.  That's what faculty members consider when they bring on people and you should, too.  Think about if you wanted a grad student - who would you choose?

 

I definitely agree with the bigger picture argument. I think I interpreted the advisor's (and original poster's) request differently. When I read the original post, I read this as a request from the advisor to find someone who would be a helpful RA, not to find someone who would be a good graduate student for the program. I don't know how big the stack of applications that the advisor gave to the OP, but it sounded like, since the OP was not invited to be part of the department's admissions committee, there might have been some other vetting going on and the OP's task is simply to pick out the top 3 for the specific task of data collection.

 

But after reading your post, it actually might sound like the advisor is asking the graduate student to help the advisor pick new graduate students for the lab/department (i.e. prior to any other vetting by other faculty members). In that case, I definitely agree with the bigger picture argument, and that the OP should take the opportunity to "think like a faculty member" and pick the best candidates for the graduate program. To me, this was not my initial thought because in my field, when graduate students are involved in admission decisions, we actually sit on admissions committees and have the full discussion with the other faculty about the department goals/priorities for this year, etc. I think it would be extremely hard to pick out the "top 3" from a stack of applications without involvement and discussion with other members of the admissions committee. 

[Crucial BBQ]

Hi guys,

 

Today my advisor gave me a stack of applicants for our lab and asked me to select my top 3 (we will hire one phd student to work with me in collecting data for my/our research). Have you guys done this before? I don't know how to choose them...

 

There's one applicant that I really like but his SOP is soooooooooo bad. Since I'll need someone who writes well, I don't know what to think about this applicant.

 

Any advice?

Are you looking for a post-doc?  I mean, you write your lab will hire "one Ph.D.".  

 

Since you are looking for someone with a particular skill set, I think the advice should be obvious.  If you are looking for someone to help collect data for your research and this person needs to write well also, then select the top three who match closest to what you need and who can write well (SOP).  Think of it this way; if your team needs a wide-receiver you can dismiss all of the point-guards amongst the applicant pool.