PIs Working At The Bench

Apr 21 2008 Published by under Careerism, Mentoring

Dr. Jekyll and Mrs. Hyde (or, as I like to call her, Dr. J) has a post up about the very interesting topic of PIs who continue to work at the bench physically performing experiments in their own laboratories. She was very impressed by the time-management skills and dedication to benchwork that this requires:

[A]ny tenured PI still doing experiments probably has some seriously efficient work habits, in addition to a deep-seated love for the benchwork. Hats off to you, Unnamed PI.

Needless, to say, PhysioProf has some opinions about this topic.


Before we get to my opinion, I also want to share with you Dr. J's good explanation of some of the benefits of PIs performing experiments, that go beyond the performing of those experiments:

Since he's down in the lab doing experiments, he's readily accessible to help troubleshoot other folks' experimental woes, and to help recognize when someone's seeing something Rilly Cool (I'm sure PP had a post about this, but can't find it, sorry). Seems like an obvious point, but again I know a lot of PIs who hole up in their offices to write grants (understandable....) and as a result miss out on times where they could easily give a boost to a lab-member's science.

I haven't physically performed an experiment since the first year after I started my lab. However, I do not just sit in my office with the door closed on my computer all day when I am in town. Rather, I make the rounds of the lab at least once a day, checking in on people's experiments, seeing what is going on, helping with troubleshooting, etc.
And I spend explicitly bracketed time on a regular basis with each of the small project-based teams within my lab discussing design and planning of experiments, analysis of data, construction of data figures and manuscripts, troubleshooting, and long range project planning and time management.
If a PI likes to physically perform experiments, and she can find the time to do so, then by all means go for it. But it is a total crock that it's some sort of moral victory to do so, and some sort of moral failing for a PI to not perform experiments with her own hands. (And I know this is not what Dr. J is implying with her post, but this pernicious idea is out there.)
Frankly, I am no more capable of physically performing experiments with my own hands than any of the trainees in my lab, and by now, probably less. What I am more capable of doing is writing grants that keep the lab well-funded so trainees can, you know, get paid and eat and have a place to sleep and shit. And I am more capable of mentoring my trainees and guiding their science in the ways I described above than anyone else in my lab. And I am more capable of getting invited to deliver seminars at other institutions where I sell the work of my trainees as being exciting, valuable, and scientifically important so that they attract attention when they, you know, are looking for their own independent positions.
Based on all this, it would be a disservice to my trainees, my lab as a whole, and our scientific progress for me to spend time sitting at the bench doing experiments. It is vastly more productive for me to spend my time on high-leverage tasks like working on grants and manuscripts, mentoring my trainees and guiding their science, and traveling to deliver seminars and invited conference presentations to keep my trainees' work prominent in our field.

22 responses so far

  • Orac says:

    If a PI likes to physically perform experiments, and she can find the time to do so, then by all means go for it. But it is a total crock that it's some sort of moral victory to do so, and some sort of moral failing for a PI to not perform experiments with her own hands

    I actually managed to keep performing experiments myself until about a couple of years ago, and this is in addition to my clinical duties. In my first three years or so as a PI I did a lot of experiments. The reason, however, was that I didn't have a postdoc and only had one technician, who, alas, needed a fair amount of supervision. If the work was to get done, I had to do a lot of it. Now I have three people working in my lab, one of whom is a PhD, and it just isn't as necessary. That's why I slowly drifted away from doing benchwork myself, and the last time I did any was probably a year or two ago.

  • I concur with both PP and Orac - I did expts myself when I had to during my first couple of years simply because I lacked hands in the lab. I also agree that, once you have one or two really good people, one's time is far better spent on grant-writing and lab promotion that PP describes as long as one stays very actively engaged in the laboratory goings-on. That is far more important than actually doing the experiments yourself.
    The biggest challenge for me when starting my lab was being able to troubleshoot experiments I had not seen being done, but you get better at it by staying involved. Now, people get really scared if I come in with the intention of doing an experiment and complain that I was the one who stole their pipetmen.

  • drdrA says:

    The first couple of years I worked in the lab on and off quite a bit- mostly because of personnel issues- but also because I had better (i.e. more experienced) hands than everyone else. I hardly have time to do this anymore though... too many papers and grants to write- it's kinda like the hands ... I've got a more experienced head for the writing now ...
    I do still kind of mourn the bench work though- just because I like doing it...I haven't yet given up my bench but it's on long term loan to a student..

  • dave X says:

    This issue has comparative advantage written all over it.
    Labs need both grants and experiments.

  • bill says:

    as long as one stays very actively engaged in the laboratory goings-on
    Postdoc perspective: emphasis on the very. While I see no need for PIs to do expts with their own hands, I do think it helps keep them grounded in the reality of lab work. I've known a lot who have lost perspective. This mainly manifests in the "when I was a postdoc I got so much more done than you slackers" pathology and the "oh, you can just finish that mountain of tissue culture in the morning and genotype four hundred mice in the afternoon" problem.
    If I make it up the foodchain, I sorta plan to try what juniorprof describes: keep a back-burner project for myself, just so I can get the gloves on now and then. (I can already hear PP and co. telling me, yeah, good luck with finding time for that.)

  • bayman says:

    This issue has comparative advantage written all over it.
    This raises the issue of the role of specialization in science. Clearly increasing specialization maximizes the efficiency of production. The argument above makes it clear how to best leverage the different specializations available to the research lab - get the PI more or less out of the lab.
    I'm wondering whether free market principles should dominate science, or whether things other than maximizing the efficiency of production might also be important. Might there be qualitative advantages, for example, in having PIs who write grants, advertise their work, but also spend some time working in the lab? Would continuing lab work influence the content of a PI's grants, their thinking on grant or paper reviews, or how they communicate their work? In other words, would production proceed in a different direction (even if at a reduced rate)?

  • DrugMonkey says:

    I can already hear PP and co. telling me, yeah, good luck with finding time for that.
    I often fantasize about taking a sabbatical in my own lab.

  • DrA says:

    Don't disagree, but just want to point out that not all science is done in the lab. But this PI/Full professor still works in the field, although the ground is lower than it used to be 35 years ago. Best excuse in the world for not being on campus.

  • bayman says:

    Ahh. Here's the passage I was looking for. This is Sun Tzu on terrain from The Art of War. It's entertaining to try and fit the metaphor to research:
    "During the Warring States when Wu Ch

  • PhysioProf says:

    Actually, PP I find it interesting that as someone who has primarily been trained to do laboratory research, you feel you are most effectively using your skills by writing grants and so forth. Maybe it would be just as or more efficient for you to train students to write the grants while you leverage your vast experience where it is more valuable - in the lab? Maybe even better would to employ specialist grant writers?

    No one gets explicitly trained to be a PI. However, some of us turn out to be quite good at it. Now that I am aware of this about myself, I exploit it as much as possible.

  • DrugMonkey says:

    Actually, PP I find it interesting that as someone who has primarily been trained to do laboratory research, you feel you are most effectively using your skills by writing grants and so forth.
    I think you have this a little mixed up, bayman. The well trained scientist is, and should be, much more than a competent experimentalist toiling at the bench by the time they are ready for an independent research career leading a laboratory. So what someone like PP has been "primarily" trained for is exactly the job of the PI. A generalist to be sure; someone who is at home at the bench, writing or presenting. Seeking and administering funding. Because at various times one may need to rely on all of those skills-especially in times of tight funding, a PI may indeed need to return to the bench. May not be able to afford to hire effective grant writers (whoever they are) or good postdocs. In times of excess fortune a PI may be forced to be "bigger" than he or she really intends.
    Without doubt, as ScienceWoman noted, one's personal circumstances and preferences alter the balance a given PI might select. But one needs to be realistic about one's goals and how to accomplish them. And some things come first. No funding, no bench work to do, right?

  • DrugMonkey says:

    However, some of us turn out to be quite good at it. Now that I am aware of this about myself, I exploit it as much as possible.
    This is pretty funny because I think (fairly seriously at times) that one of the things I am "good at" is taking the punishment of grant rejection, getting up off the floor and revising that sucker.
    One thing that is pretty interesting to me is that I know at least two people that in the postdoc stage it was clear that they were best-suited to be a PI. In one case the postdoc was fairly universally recognized to have bad hands, had attentional problems that were incompatible with his type of research etc but was a great big-picture guy. He's now a PI at a high-falutin' institution and by all appearances doing great. Good enough supervision of trainees combined with lots of international schmoozing equals success. The other postdoc was just basically too lazy to put up with the drudgery expected in his subfield of face-time lab work. Bright guy, lots of big picture ideas and I have no doubt whatsoever that he would have been a successful PI. Instead he went into an job that requires a lot of analysis of science, big picture understanding and communicating "teh excitementz" of science, in short expressing skills that would have been great PI skills if he'd managed to navigate the process into PIdom.

  • TeaHag says:

    The hands/brains issue sometimes forces PIs into the lab (along with the limited funds issue). I work on a rather complicated intracellular microorganism and from time to time there's nothing for it but to go in there and do the experiment myself... after all, I've almost 12 years intensive experience just looking at the darned thing... you notice stuff and file it in your brain.
    Still, I remember from my post-doc days the self-indulgent crap that some PI's engage in with the "I'll do the experiment myself"... turns out to mean that they set up the PCR rxn using the primers you designed and the template you prepared, and then rely on you to run it out/purify/clone/confirm construction etc. Then you have to listen to them prat on about how they "love to keep their hands in". So I never start anything that I can't finish and I've promised my post-docs not to ask them to CMA!!

  • Becca says:

    Bravo to bayman and bill for touching a critical point.
    The major reason I appreciate PIs at the bench is that it can help remind them how freakin hard and/or frustrating the methods can be.
    In people with average, or even good (maybe not great) 'hands', nothing works every time. Not a Western, not a PCR, not even making up solutions. It doesn't matter how many times you've done it (though experience can often tell you if you've made a small error or a fatal one). It doesn't matter how "easy" it is. It doesn't matter if have "an established protocol".
    My humble suggestion for PIs- do what you're good at and enjoy. But be extraordinarily cautious when describing techniques as "trivial"- particularly when your lab members have never seen you at the bench.

  • bill says:

    I think (fairly seriously at times) that one of the things I am "good at" is taking the punishment of grant rejection, getting up off the floor and revising that sucker
    This is crucial to more than grant applications. Doing research is in many ways largely a test of your ability to keep going despite constant setbacks. It's just the nature of the game: most of the things you try will fail, most of your ideas will turn out to be crap; and when I say "most" I mean 90%+.
    A capricious universe, utterly indifferent to your misery, is going to kick you in the metaphorical nuts on a daily basis, and if you can't keep picking yourself up and moving forward you're not going to get anywhere.
    I once saw an NIH applicant review form -- as in, you fill this out as a reviewer for a job applicant -- which ranked "ability to deal with frustration", those exact words, as one of the most important factors for the referee to comment on. I realized then and there what much of my job time was going to entail, and I've been trying ever since continually to improve my response to setbacks. Along the way (I like to think that) I've picked up a pretty good line in coping with criticism, whether well meant or not.

  • Agreed with Becca on the "be cautious when describing techniques as trivial" bit. Nothing makes a grad student feel unappreciated like having her hours of troubleshooting and frustration described as trivial....I've seen this happen and it leads to a lot of resentment and muttered comments about "wouldn't know a pipetman if one was shoved up his ass" or similar.

  • bsci says:

    While I don't see a reason for PIs to do benchwork with the goal and time commitment required for publishing ones own data, I've seen a major flaw in some PIs who only devote their time to grants and mentorship. In addition to losing some memory on how to do specific tasks, they fall behind on the practical steps of doing something new. Reading a paper is nothing like actually doing the work. It's one thing if you are reading others papers, but it's another if you are reading your own papers.
    One thing I've seen happen is the game of telephone where each student/postdoc teaches a technique to others and assumptions and flaws in a technique get passed along. These are often at a lower level than what might come up in mentoring sessions beacuse they've become standard lab protocol.
    If a PI just went through the full data analysis process even once a year, she'd be better able to keep track of what's happening and be a significant piece of institutional memory.
    I've also seen lab knowledge gets too concentrated in a few individuals and sometimes the PI is so far removed from the bench that the value of these people are not understood until they leave the lab (no I'm not talking about myself).
    There are solutions for both of these besides doing some token benchwork, but benchwork is probably clearly way to take charge of knowledge transfer and the buck stops with the PI.

  • ecogeofemme says:

    The PIs I work with often lament that they don't *get* to work at the bench or in the field anymore. They spent all that time learning to do the hands-on work and now they sit at their computers in their offices all day. While I'm looking forward to graduating and developing a research program, I will be sad when I don't have time to get my hands dirty anymore.

  • bayman says:

    DM said,
    A generalist to be sure; someone who is at home at the bench, writing or presenting. Seeking and administering funding.
    Agreed - the PI should try to find a balance these things to have a maximally effective lab. My point was not that the PI should only do lab work, but that there are many more benefits to having PI spend some time working in the lab than simply the extra set of hands. In fact, many of these benefit could even be gained just by getting in the lab a few hours every few days and pipetting water back and forth.
    I have seen PIs who can't get a grant to save their lives (literally) simply because they don't have the faintest clue what and how things are being done in their own labs. And they have no perception whatsoever why they're failing to get meaningful data because they live in the insular world of PI-office land, where "established" PI colleagues convince them that their time is best used only at the keyboard and not at the bench.
    If I was a PI I would also be concerned about the wide culture gap that develops between PIs and their underlings when none of their working life is spent in the same environment. This can make it nearly impossible to understand one another and communicate effectively.

  • PhysioProf says:

    Bayman, your concerns about PIs losing all contact with the experimental reality in their labs are very salient. However, the solution is not for PIs to spend a few hours a week at the bench pipetting water back and forth. The solution is for PIs to spend many hours a week in discussions with their trainees about experiments.

  • Dr Aust says:

    In non-off-the-shelf apparatus-heavy kinds of science (like a lot of physiology, at least pre the "molecular revolution") the PI also tends to be the person who built the patch clamp rig (or similar) and can troubleshoot and fix the equipment. That is certainly my own sole "hands-on" contribution to the lab apart from training people up to use the rigs (easier to teach how it works if you put it together).
    I'm agnostic about the "actually doing experiments" thing. Bayman's comments are spot-on, as are Becca's about PIs forgetting just what a !*!!ing pain it is getting lab science to actually work, something that I think gets very marked in senior (full Prof) PIs. One can go "kibitzing and discussing", and it works well for some folk. Others need the direct "discipline" of doing it themselves. One thing I have seen work is people who do preliminary / tryout / "what if..." / mad idea stuff themselves, while the lab grunts do the bread and butter things. But anyway, it's largely a "what works for you" kind of gig.
    Historically many older scientists got into it (thinking back 20+ years) out of enjoying DOING science, rather than enjoying organising other people to do it, so those of my generation (mid 40s up) tend to admire the "still go in once in a while and do it themselves" types, even if we're not like that ourselves.
    An interesting aspect of this kind of PI is that, even when pretty famous, they don't tend to run big labs (

Leave a Reply