Tuesday, July 13, 2010

Helping students become scientists

One of the most difficult tasks I am finding as a new TT professor is in teaching my students how to be scientists. Teaching them the nuts and bolts of how to do experiments--no problem. Showing them how to do a literature search--piece of cake. Helping them prepare talks and posters--more of a challenge, but I am up to it. Getting them to think like scientists, and not just people who can repeat protocols accurately--this is the hard part.

My group has reached the point where our lab is (mostly) set up and everyone is making some progress on their projects. Meetings are much more fun now, with new data to talk about nearly all the time. One of my grad students is doing extremely well--she is thinking up new experiments, figuring out how to test out ideas, bringing in new literature to think about, and generally owning her project. Some of this is of course very raw, but I can see how she is starting to think scientifically and creatively.

My other grad student is making good progress on data acquisition (which is of course important), but still needs a lot more guidance. He only does experiments we have discussed in detail before hand, and never does any kind of follow up (beyond rudimentary data analysis) on his own. He is happy to go ahead and do more, but only after I tell him what to do next. I find myself sending him papers on his project (which is normal), but he never looks to see how they have been cited, unless directed to specifically (which I've done many times). He isn't lazy by any means, just not very independent. I know he has only been in the group a year, but my summer students seem more independent (if less experienced) than he does.

I am thinking about my approach to mentoring--I have a fairly hands off approach that seems to be working well with 4 of my 5 lab peeps. I talk to everyone at least once a day informally, and walk through the lab a few times in case people want to talk. Maybe this student needs some more structure? Some more limited independent projects to try on before tackling his PhD project? I know that not all approaches work for all people, and I don't want this student to fall through the cracks through my own failure, since he is smart, interested in science, and a hard worker. My own experiences are not a good guide, since both my PhD advisor and my postdoc advisor were extremely hands off (more even than I am!).


GMP said...

Some people need a lot of hand holding. I don't think they ever make academics, but if they are hard working they can be useful members of a team. The student you mention can probably be coached into finishing a PhD but make sure he explores career options where he would be a follower in a large team rather than a leader.

Hope said...

Some grad students do need a lot of hand-holding at the beginning, but I disagree that this means they’re not destined to become academics or research leaders. How ‘bout giving the kid a little more time before you decide he just doesn’t have what it takes?

Have you talked to him and told him that you’d like to see him take some initiative more often? Especially if he’s not from the US, he may need to hear your expectations made explicit. To some people, it’s not obvious that it’s better to ask for forgiveness than for permission. But that doesn’t necessarily mean that they’re not cut out for independent research.

prodigal academic said...

Thanks for the comments.

GMP, I think he definitely would be a good industrial scientist. He works well with others, and is a hard worker. He doesn't yet know what he wants to do with his PhD, which might be part of the problem. I just don't want him to regret getting he is coached through--the one person I know who is still working in science who regrets their PhD really wants to be an advanced technician, but is no longer considered for such jobs due to the PhD.

Hope, I don't plan on writing off the student at all--in fact, I am struggling now with how to change our interactions to help him reach the potential I see in him, since i have been unsuccessful thus far. He is from the US, but that doesn't mean that cultural issues aren't present. I will definitely try a more direct approach. His research progress has been good, it is just the other parts of being a scientist that are lagging.

Dr.Girlfriend said...

I think the best way to train him is to start treating him like a scientist. Lab meeting help a lot with critical thinking, trouble-shooting, and hypothesizing. Ask him questions rather than guide him. Get him used to the kind of questions his committee will ask at his defense.

Have him regularly present his immediate and long term research plans to the group - it will force him to plan, answer questions from the group, and defend or rethink his ideas. Seeing others go through this process will also help he get a better idea what a scientist is.

The difference between a technician and a scientist is that the latter should able to develop new techniques, trouble-shoot, and critically analysis the quality and meaning of the data.

Listening to senior scientists talk science, rather than present science, will help him understand process better - joint lab meetings, department journal clubs, or small conferences are great places for this.

It might well be that science is not for him, but let him figure that out.

melissa's said...

I am in a very similar situation, with a very similar mentoring style and one student that requires additional hand-holding. I'm happy to give it, and it's still early days, but I wanted to add one thing I've sensed. This student was a straight-A, hardworking student at a top undergrad institution, and I find his ingrained undergrad mindset is part of the problem. He seems unsure of himself unless given assignments with explicit instructions. A corollary to this is that he often tries to troubleshoot on his own, despite the ready and friendly presence of myself and other labmates, leading to unnecessary wasted time and mistakes. The whole informal "let me bounce some ideas/problems off you and see what we can think of as a solution" thing is foreign to him. It will take some time to teach him that, in research, he is not an island like he was in college, expected to produce results without help.

Bob Carpenter said...

I think Melissa's hit the nail on the head here. Lots of students have a mistaken notion that everything has to be done independently as if they were undergrads taking a test.

I used to let students get help from whomever they wanted on their homework as long as they cited the helper's contribution. Surprisingly, only the very good students took me up on that offer. I partly wanted to encourage them to think like they were in the real world, where you can get help from whoever'll give it to you.

How much useful work can be gotten out of a non-independent (and even an independent) student depends a lot on the prof. I was never good at getting work out of non-independent students because I'm a terrible manager -- I barely know what I want to do ahead of time. A former colleague of mine who's still at Carnegie Mellon, Lori Levin, was (and presumably still is) a genius at getting useful work out of just about anyone.

Dr.Girlfriend said...

Ref the last couple of comments -

I have to confess I did not go to grad school because I wanted to become a scientist. I went because I loved to take classes, and wanted the highest degree obtainable. The research part was simple an agreeable way to earn a living.

For the first 3 yrs, taking classes was my first love and priority. I did what I had to in the lab, but I did not really sink my teeth into until I was done with classes. I only buckled down then because I had to if I wanted that coveted PhD.

For me it was the technicalities of doing experiments that I struggled with. I could write proposals, but hands-on I lacked the patience.

I think if I had been pushed to write and present my research more during the first few years I would have been more inspired to get the tedious benchwork done.

Hope said...

@ PA: I didn’t mean to imply that you were going to write your student off. My first paragraph was more a response to GMP’s comment than to your post.

I think it’s great that you’re looking for other ways of getting through to your student, instead of saying, “well, this is what I got, so if he can’t make it on that, then too bad.” Understanding that students have different needs and goals, and being willing to adapt your style to do right by all of them is what separates the great advisors from the merely OK.

GMP said...

Hope, I am not dismissing anyone. Prodigal Academic knows how the student is and I am sure she will do what's best.

I am sorry, but I stand by my opinion that, with experience, you can tell fairly quickly what a students' skills and potential are. I didn't say the student is to be dismissed, but that a student who constantly needs hand holding is not likely to make an academic.

There are international students who initially struggle with adjusting, but I assure you that talented ones will come out unambiguously if you know what you are looking for, even if they are struggling with language and a culture shock. It is very hard to hide talent, or show it if it's not there.

In the past, when I would find a student struggling, I would keep making excuses for them -- they need more time, if only we changed the project one more time, if only I found yet another way to approach what they need, yet another way to stimulate them.
I am sure this won't make me popular with you, but every single time the assessment of the student's potential I had after the first 6-9 months turned out to be accurate even after 2-3 years and much more money and effort spent. So now I completely trust my gut about the student's potential; this is not discriminatory, it actually helps quickly devise a way how to best mentor them when you know what you are working with.

I agree that different students need to be mentored differently, but I would like you to consider two points of mine: (a) that it is in fact possible to assess a person's potential fairly quickly, and (b) that now everyone is meant to be an independent academic. I think people who need constant hand holding as PhD students can achieve great technical prowess, but in my experience do not make successful academics. I am sorry, but what PA described -- a student who needs a lot of micromanaging/supervision after a year in a lab
does not sound like a budding academic. That does not mean that he is not a good and hardworking PhD student, and can do well with proper guidance which I am sure PA will provide.

prodigal academic said...

Melissa and Bob thanks for the insight. That is something I hadn't considered--this may be the first time my student hasn't been able to do everything himself, and that may be part of the problem. He is as you describe--a straight A undergrad having trouble adjusting.

Dr. G--I think that motivation is part of the problem. My student does not know what he wants, just that he really likes science, and is a good student. He may be finding it hard to self-motivate the way a PhD needs to without a direction in his life.

GMP, I appreciate your words of experience. I will do my best for all of my students, but I totally agree with you that not everyone is cut out for a career as an independent researcher, and also that there are plenty of places such a person with a PhD can thrive. The two pieces of being on the TT that I am learning from scratch are how to evaluate potential lab members and how to mentor them (when hands off doesn't work), so I definitely appreciate the comments.

Thanks to everyone for the nice discussion. This is why I really like having a blog!

Hope said...

GMP, I’m very glad that this student is in PA’s lab and not yours.

I have two thought that I would like you to ponder:

a) Do you share your 6-9 mo. assessment of your student’s potential with that student? If my advisor had concluded that I didn’t have what it takes to become an independent researcher, I would want to know … so that I could find myself another advisor. I prefer to be mentored by someone who believes in me and my abilities.

b) Have you considered the possibility that your “assessments” may turn into self-fulfilling prophecies, since they apparently influence how you mentor the student?

And by the way folks, independent researcher != academic. I agree that not everyone should become an academic, but people that are getting PhD’s are, in general, not doing it to become technicians.

I always thought that the whole point of getting a PhD was to learn to do independent research. If you can tell at 6 mo. that someone will never be able to do this, why keep them around for 6 yrs?

GMP said...


For someone who is worried about students being dismissed too soon, you certainly didn't take very long to completely dismiss me as an advisor!

Not all people do a PhD to be able to do independent research later. I actually ask my students what they want to get out of the PhD experience: some don't know, some want to become faculty/independent researchers, and a large portion want to obtain some marketable skills and an advanced degree and go get a well-paid job in an industrial team. I am not sure what your first year on the PhD was, but I assign my newbie students a series of mini-projects to test their technical preparation, resourcefullness, initiative, creativity, work habits, and after these are completed we meet again -- usually several times -- to decide what type of project is best suited for the student. So yes, after 6-9 months, both the student and I have a very good idea what types of projects he/she would be most successful at doing. Sometimes, the student turns out to be a poor fit for the group altogether, and it's best to cut them loose sooner rather than later so they can find another advisor before too much time is lost.

(Prodigal Academic, my apologies for hijacking your comments with this exchange.)

Hope said...

Not all people do a PhD to be able to do independent research later. I actually ask my students what they want to get out of the PhD experience … and a large portion want to obtain some marketable skills and an advanced degree and go get a well-paid job in an industrial team.

GMP … I totally agree with you that not everyone is cut out for a career as an independent researcher, and also that there are plenty of places such a person with a PhD can thrive.

GMP and PA, in my area (physics/engineering), when industry shells out the big bucks for a PhD, they expect someone who can do independent research, and often, someone who can direct the research of others. Ditto for government labs. If a student just wants a spot doing research on an industrial team, they are much better off applying with an MS. This is what my 10+ yrs of work experience outside of Academia has shown me. So I’d be very interested to know where you think that a PhD incapable of doing independent research would thrive. Perhaps that could be the topic of a separate post.

GMP said...


A large number fo my colleagues from grad school went to work for Intel. Only 2 of them work in R&D and do what could fall under independent research (although I would say they work as part of a research team, with a string of supervisors above them). The others are all PhD's working on reliability, testing and failure analysis, etc for the more mature technology nodes. So well-paid highly specialized jobs for PhD's (not MS graduates).

prodigal academic said...

At least half of the people in my program at PhD U always intended on a career in industry (and not necessarily in research). When I was interviewing when I finished up, I saw many non-research PhD level positions that require advanced skills, a thorough understanding of the field, the ability to learn rapidly and explain things to others, but don't involve research, as GMP mentioned above. This is a great place for someone with PhD level science skills but poor abilities at independent research.

At National Lab, I knew several people who did not do independent research. They worked on other people's projects, but never wrote proposals. This job path is a lot harder to get on now (the way things are changing at National Lab), but such people are HIGHLY sought after by PIs, because they are often fantastic in the lab, just not so good at/so interested in the big picture stuff.