How to assess ability to look things up for yourself in an interview

February 6, 2015 9:29 AM

What question can I ask undergrad research assistant candidates to see whether they are good at looking things up on their own?

I'm a senior grad student in the sciences managing a project where I need a bunch of undergraduate research assistants to do fairly clearly specified but not mindless work.

I've noticed in working with undergrads (and, to be fair, everyone else!) that they vary widely in their ability and willingness to find information on their own instead of waiting to ask me. Obviously sometimes I actually am the best source of information ("Am I on the IRB protocol yet? Am I allowed to view videos marked XYZ?") but other times I am clearly not ("How can I hide a column in Excel? What should I do in the event that [situation clearly spelled out in lab manual] happens?"). I suspect this may also correlate with a tendency to THINK about things on their own, hypothesize explanations/solutions, actually try to debug code instead of deciding "it just doesn't work," etc., and am open to comments on whether this is plausible.

I'm not so much concerned with minimizing the frequency with which I get asked Google-able (or lab-manual-able, etc.) questions due to irritation as I am hoping to find students who take more responsibility for learning or finding information on their own.

How can I get a correlate of this ability during the selection process? This process currently consists of an in-person interview with a discussion of experience and a few reasoning questions (not on par with Google's), plus one informal reference check. I'm open to anything I could ask students to do during the interview (e.g., what Google-able question would be informative to have them answer?) or anything else I could add to the process (e.g., a question posted in the job listing or emailed when they apply).

I can afford to be fairly selective here (generally taking 1-2 students out of 5-10 applicants, from a top-several school in the sciences), so questions that could identify exceptional students rather than a "low bar" they should absolutely be able to pass would be great.

posted by cogitron to Education (13 answers total)

"Tell me about a time that you took the initiative to find a piece of information that wasn't readily available in your textbook or class notes."

"Tell me about a time that you had to go back to a textbook from a previous semester."

I find these kinds of prompts kind of awkward to deliver in an interview, but if you ask about a skill that you really need, but place it in the context of their previous experience, the answers can be quite illuminating. They will go all the way from the candidate giving an exceptional example of exactly what you want them to be able to do, to "Uhhhhhhhh, no I can't really think of any time I did that."
posted by BrashTech at 9:55 AM on February 6, 2015 [6 favorites]

I would try coming up with different hypotheticals, and asking what approach they'd take to solving each.
posted by mmiddle at 9:57 AM on February 6, 2015

"Describe a time you had difficulty on a homework problem. What steps did you take to deal with this?"

I would immediately discard anyone who said "I've never had any difficulties", who complained about the professor or TA, or anyone for whom no part of the answer involved doing any legwork on their own.
posted by tchemgrrl at 9:57 AM on February 6, 2015 [1 favorite]

"Tell me what steps you'd take to find X."

You should hear about process and lots of ideas.
posted by Ruthless Bunny at 10:03 AM on February 6, 2015 [3 favorites]

Best answer: I would make sure you're also figuring out a way to screen for people who research and do things on their own that they really should have run by someone with more knowledge instead.

There's generally a balance between "takes initiative" and "overreaches because they don't know what they don't know."

I don't have good questions for sussing out people who fall into the productive part of that scale, but I suspect that the "Tell me what steps you'd take..." type questions would be most helpful.
posted by jaguar at 10:10 AM on February 6, 2015 [5 favorites]

Since you said you're open to whether it's plausible, I'm going to say "no." I mentor undergraduates regularly, and I still have no idea how to predict which ones will be able to do this and which ones won't. I understand how annoying it can be, but the opposite problem is equally bad - thinking they know how to solve something or figure it out and then spending a ton of time doing the wrong thing without checking in. Knowing when to ask your advisor/mentor for help is exactly the kind of skill they should be looking to learn from you when they sign up for a research experience. It's true that some people just have it naturally, but it's not like an innate talent that can't be learned. Conversely, teaching them that ability is a mentoring skill that you can pick up (and is a great, specific thing to mention in cover letters, interviews, etc.). I find that making this explicit in the interview is the most important part: "Part of you becoming an independent researcher is learning how to help yourself. We're going to be working on when it's appropriate to ask me for help and when you should try to figure something out on your own." And then practice what you preach. If they come to you with a question that you think they should have solved on their own, ask them what steps they've already taken before coming to you. Then, show them what you would do to figure out the answer.

On the other hand, when I first started confronting this problem my wife had a simple solution. Invite them into an interview where you're putting a puzzle together. Tell them the puzzle is time sensitive, and then get called out of the room. The ones who put it together while you're gone are probably the ones you're looking for. Never tried it, couldn't come up with a time-sensitive puzzle.
posted by one_bean at 10:14 AM on February 6, 2015 [10 favorites]

I typically use a Google based process to teach myself how to use new software or hack together some html or whatever. I wonder if asking about how they learn to use new software might give you some insight into how they learn to do things generally.
posted by papayaninja at 10:24 AM on February 6, 2015 [1 favorite]

Best answer: In my industry, if you were applying for a corporate job, you do have to take a test that you often complete online and submit at the end. Applicants are rejected/or go high up on the hiring list based on how they do (and even though it might involve reviewing a few journal articles or finding peer reviewed articles, people fail).

Anyhow, I would write out a few of your questions, design it to take 30 minutes, and give the undergrads an hour on the computer to complete it. It might be "do X in excel" (and save the document to show you), find 10 references about this previously completed experiment (and they can cut and copy the list from pubmed - although this might be a stretch for an undergrad ), or whatever it is that you think that they should be able to reason their way through. Ask them to explain how they figured out their answers at a later time point.

But you can review all submitted tests at a later time point and quickly figure out - no attempt, wrote something fluffy with no info, to able to answer most of them and figured out how to find answers independently on their own. But I think that this would be better than a question in person because people can feed you what you want to hear (vs trying this and getting results on their own).

Also, during the interview, I would stress (with the candidates who do well) that this is a skill that you value. Because at least in my experience, some undergrads won't do this even though they have the ability.

I don't know if you have had the chance to TA (or know a grad student who TAs a lab course) because if there are some undergrads in lab sections who are independent/motivated/put in the time on their own (and it pays off) and that is the student that I would grab in a second.
posted by Wolfster at 10:26 AM on February 6, 2015 [3 favorites]

"If you run into something unexpected at work, tell me in as much detail as possible what your thought process would be as you decide how to deal with it."

Give them one freebie opportunity to ask you for clarification of the question, but don't clarify. Explain that you are being intentionally vague to try to understand how they would work through uncertainty.

The low bar would be that you don't want to hear, "if my supervisor has already told me what to do in that situation, do it. If not, ask."

Ideally, you want exceptional candidates to think through a very detailed flowchart, something like "is it an emergency? is it a non-emergency safety issue? can I look it up in the manual? can I look it up on Google? if I should ask my supervisor, does it have to be now, or can it wait?"
posted by Bentobox Humperdinck at 12:15 PM on February 6, 2015

Best answer: I don't think students would openly admit that they prefer to ask others for help rather than find information independently. I agree with Wolfster, and wonder if you can also just quickly check the browser's history immediately after the task to see what they did (how many pages deep into a Google search did they go; were their search terms well-formed; did they use YouTube tutorials or software companies' help pages, etc.).
posted by cotton dress sock at 12:27 PM on February 6, 2015 [1 favorite]

Hold the interview in a relatively-unknown, but Google-able, location on campus.
posted by melissasaurus at 12:41 PM on February 6, 2015 [4 favorites]

It sounds like the quality you're looking for is resourcefulness.

Actually I think Brashtech's questions are not awkward at all but are totally appropriate in the context of behavioral interviewing (basically questions regarding prior behavior predicated on the idea that it predicts future behavior).

If you're working in a lab and you know they've had at least a little experience can you ask: Tell me about a technical problem you've had in your labwork. Tell me your process for how you tried to resolve it?
posted by noonday at 12:45 PM on February 6, 2015

Why not just lay it out as SOP. I know this seems unrelated, but actually apt. rental leases can be good about this. Rather than get overburdened with tenant calls, the landlord may state in the lease: "These are the things you can bug the landlord about:... These are the things you need to take care of on your own:..."

So once you find a top RA, lay it out: "These are the kinds of things I expect you to figure out on your own and these are the kinds of things I would expect you to come to me about. Do any other examples come to mine. What about this scenario...? And this one...?" Some people mean well and are hard workers but in their quest to do a good job ask too many questions. They respond well to being told not to.

Or you could do it in the interview: In your experience in the lab, what kinds of things is it important to go to your supervisor about? What kinds of things did you decide would make you more productive if you figured it out on your own, through a Google search or asking another RA, for instance?" That also primes them for what you'll expect if they get the job.
posted by lillian.elmtree at 2:00 PM on February 6, 2015