The "right" size for a lab

Sep 22 2011 Published by under [Education&Careers]

From time to time the topic of lab size comes up in the blog-o-verse, with as many opinions as readers. I've always viewed lab size as highly field-dependent, with both culture and funding levels playing significant roles. For that reason, I was interested on a passage I came across in a text book I am using for a class I am teaching.

The book "Scientific Integrity" is written by Francis Macrina, former Director of the Philips Institute of Oral and Craniofacial Molecular Biology at VCU and current VP of Research. Dr. Macrina is no stranger to NIH funding, with 10 R01s in RePORTER, as well as a number of other funded R-mechs and a P30. Page 44 (third edition) of the text falls in the chapter on "Mentoring", and Dr. Macrina writes the following advice for graduate students seeking a mentor:

There is a point of diminishing returns in the number of trainees who can be effectively mentored. When that threshold is crossed, the ability to responsibly guide trainees is compromised and the viability of the training experience is put in jeapordy

Notice that Dr. Macrina puts no "threshold value" on the table, but again, that is going to depend on many factors - not the least of which are the PI's ability to juggle responsibilities, interest in mentoring and lab training structure. One PI might be ill-advised to take on a dozen graduate students and no postdocs, for instance. But I do think the question of how effective a mentor can be as the lab group grows is an interesting point to consider.

The argument I often hear from those advocating large lab groups often revolves around establishing an effective lab hierarchy as a means to spread the mentoring burden and essentially an "economy of scale" approach to working on difficult problems. The first point embraces the "business model" of running a lab, where the pecking order determines access to the PI's time. Junior grad students report to senior ones, who report to postdocs, etc. The second point posits that by concentrating resources in one place, you maximize the potential of a lab to accomplish great things.

For discussion purposes, I'm more interested in the first point - mentoring in a large group. I don't think anyone would argue against the statement that trainees in a smaller lab, generally speaking, get more direct contact with the PI than those in larger labs. I'm sure there are anecdata out there where this is not the case, but I'm comfortable with the generality here. It's simply a case of the number of mouths to feed. So the question becomes, although a lab's productivity can be managed effectively using a hierarchy, is there also good mentoring happening in the trainee/PI relationship. That is to say, does the PI have the time to fulfill the mentoring commitment that a trainee has the right to expect, or must they settle for mentoring by proxy? And is that a decent substitute?

No doubt that more senior trainees get valuable experience mentoring junior ones, but they are learning to mentor as they go*, so there may be some first-pancake-syndrome. From the junior trainee standpoint, no one wants to be a guinea pig. Would you go to the dental school for a root canal? It may also be the cases that certain trainees have little interest in developing a mentoring relationship with those at the next level down. Also bear in mind that students signed up to work with Dr. HotShot and not her postdoc.

I'm sure there are numerous strategies that mitigate the issues that can pop up as the lab grows and maybe making do with a less-than-ideal mentoring situation is as good as it gets. There will also be huge variance in the number of people it takes to tip the balance between effective and ineffective mentoring, depending on the PI. Based on the work I do, I am unlikely to ever be in a position where this becomes a major issue, but mentoring (or lack there of) is a common conversation topic among large lab dwellers and it often makes me wonder whether they are getting what they bargained for.

Alright, commence telling me how naive I am...

*Although I think PIs are often learning more about mentoring all the time, there is also the benefit of experience.

8 responses so far

  • That is to say, does the PI have the time to fulfill the mentoring commitment that a trainee has the right to expect, or must they settle for mentoring by proxy? And is that a decent substitute?

    It takes a village to raise a grad student?

  • Eli Rabett says:

    The best argument about big groups is that they produce too many PhDs who go out and chase too few grants.

  • Simjockey says:

    One that has surprised me in grad school is that even at my well-regarded R1 institution, there are quite a few professors who don't "grok" all the details of their students' work. Of course, the professors still understand the "Big Picture" better than their students, they know how to crank out the grants that keeps these students funded and they know how to present the research in a way that gets papers accepted. So, clearly these professors are doing their job, but a some of these students aren't very happy with the aspect of mentoring that involves doing science. One such professor operates a rather large research group, so what struck me about this situation was that even if this professor magically found a lot more time to give his students, it wouldn't all be useful to them.

  • ecogeofemme says:

    My grad lab was too big by the time I left. It wasn't really all that big in the grand scheme of things (<10 students), but it was too big for that particular professor to handle under those particular circumstances. Everyone resented it.

    The bigger the group gets, the better the group's organizational skills need to be. In my experience, that starts with the PI.

  • ReHoScro says:

    The PI is there to advise on where to find the deer and how many to salt for winter. Not to teach how best to remove the alimentary canal, StimJockey.

  • proflikesubstance says:

    what struck me about this situation was that even if this professor magically found a lot more time to give his students, it wouldn't all be useful to them.

    I think your definition of "useful" may be a touch on the narrow side. If your PI knows the details of your project better than you do, you're doing it rong.

  • Simjockey says:

    I think your definition of "useful" may be a touch on the narrow side. If your PI knows the details of your project better than you do, you're doing it rong.

    I see what you're saying here, and I think you're right. However, what I was referring to was not that the PI knows more about the details of the project but that the PI doesn't understand all the subtleties of the system that's being built/studied.

  • Aside from the concentration of cash and resources, the possibility of participating in a hierarchical mentoring scheme (as a mentor) is perhaps why the "training environment" score is typically high for large labs. There are clear disadvantages to being the first pancake, but I think they may be more severe if your chef is a PI rather than a senior grad student or post-doc.

Leave a Reply