All PIs Totally Suck And Only Get In The Way

May 30 2008 Published by under Conduct of Science

MsPhD has some interesting responses in the comments to my post from a few days ago addressing the issue of strategic planning of an experimental research program. Here are a few particular excerpts that I will address below the fold:

Our mentors, beg to inform you, have ZERO novel ideas of their own.

I have NEVER met or heard of a 'mentor' who knows the ins and outs of technical things as well as the lab members do.
In this day and age, there are no PIs who can keep up.

[I]t belies the Apprenticeship part of the system. To get new things to work at the bench, you have to be willing to work at it. Yourself. Your mentor will not help you.


Just to be clear, MsPhD is speaking here about her own experiences--which seem to have involved a severe lack of effective mentorship. This is very sad, and represents a failure of the system.
However, this does not mean that there are no good PIs who are highly creative, excellent methodologically, and extremely effective mentors. There are many.
We really, really, really needs to get past the canard that it is a failing of mentorship and lab leadership if a PI does not (or even cannot) sit down at the bench side by side with a trainee ("apprentice") and teach the trainee the physical process of performing a particular technique. This has nothing to do with being a good PI and an effective mentor. It bears no correlation with whether a PI is good at generating novel ideas or techniques.
I know that it is hard to see from the very limited perspective of a post-doc who spends her day at the bench physically performing experiments, but the best troubleshooter of an existing technique and the most creative inventor of new techniques can be a PI who couldn't possibly sit down and perform them herself.
Sitting at the bench or having good hands has nothing to do with being a good PI. Nothing. Nada. Zilch. There is no positive correlation.
Some of the best PIs I know were nearly hopeless at the bench as trainees (and presumably would still be if they tried to do benchwork). And some of the most ineffective PIs I know--whose trainees are left to hang out to dry with no guidance at all and who tend to run away from these labs after a few years of no productivity at all--are themselves outstanding experimentalists and spend a lot of time at the bench.
In fact, if anything, there may be some negative correlation. Those PIs who are outstanding experimentalists could be highly effective post-docs without even beginning to develop some of the skills required to be a PI: leading a group, troubleshooting experiments you didn't perform yourself, being creative without being at the bench, etc.
Those PIs who are not outstanding experimentalists had no choice as post-docs but to develop these other skills, or else fail. So they developed them. And now as PIs, surprise surprise, they are effective, while many of their "better hands" compatriots turn out to have no choice but to spend most of their time at the bench themselves doing the experiments that they cannot lead others at performing, and not only have not developed talent at mentoring, but also no time for it.
It sounds like MsPhD has experienced an unfortunate lack of good mentorship, and her perspective on the role of an effective PI has been, not surprisingly, colored heavily by this. However, although she may not know this, that lack has nothing whatsoever with whether her PI can or cannot sit down at the bench and perform an experiment. That is a total red herring.

72 responses so far

  • TreeFish says:

    I gotta agree with you, PP.
    I don't think Eric Kandel or Rick Huganir or Roger Nicoll sit at the bench, yet they have enormous and enormously successful labs (including EK's Nobel Prize and maybe one for Roger some day). Sure, when they were post docs and Asst Profs they had tremendous hands; their transition into uberscientists, though, came when they formulated a vision, incubated talent, and communicated their results with nicely written prosaic accounts.
    As far as intuition, let's look at an anecdote. Roger Nicoll and David Bredt had a wonderful collaboration before David went to Eli Lilly. In his talks, David would always poke fun at Roger because Roger had tremendous intuition but didn't understand a thing about the complicated molecular biology that David is so great at (David's imitation of Roger was head down, scrunched face, arms waving, and saying 'molecular biology stuff'). Yet, their collaboration led to huge discoveries (e.g., TARPs as auxiliary subunits for AMPARs that direct their trafficking to synapses) and some very successful youngsters (e.g., Susumu Tomita). Roger did the same thing with the inimitable and indefatigable Rob Malenka (who is in his own right great).
    Vision and insight seems to be what's needed for a PI. Bundling the vision and insight into digestible packets for trainees can sometimes be tough, but the ones who do usually win out in the end. And when they win, science wins.
    Overall, though, one consideration is that these huge labs have a lot of failures to accompany their successes. So, perhaps, a great mentor is in the eye of the beholder because even the best mentor can't bat like Ted Williams.

  • Propter Doc says:

    I can't actually believe someone could write that about a PI, generalizing so much. Of course PIs have ideas, of course some are better than others, some mentor, some run a 'survival of the fittest/smartest' lab. All different.
    The idea that being good at the bench is no real training for being a PI is one I'd have to agree with. Nothing that was in my postdoc 'job description' (i.e. bench work) prepared me for what I do now. The additional roles that I sought out during my postdoc - random bits of service and committee work, teaching here and there, being involved with non-bench work stuff, well, I'm just starting to realize how important that actually was to do. Let me be clear I'm not talking about reducing research productivity to take up a series of pointless admin tasks, but making time in schedules to move beyond just benchwork, showing an interest in life outside the lab, trying to get some experience of how wider university politics work. This also shifts your perspective as a postdoc away from the bench and into the wider university community if you aspire to be a PI. It was about the only thing that kept me sane over the last few years.
    (and as an aside, I'm one of those people who can deal with lab work, but will never be listed as fantastic at the bench).

  • matt says:

    I guess I just got lucky then. Great grad school PI, great post-doc PI and I hope to be a great PI when I start my own lab next year. Stay out of the labs with 20 post-docs, stay out of the labs where the PI is super-huge and is never around the lab, stay out of a lab where the people in the lab are not excited about their work, and stay out of a lab that isnt asking an interesting question.
    If you are excited about the work, if the people around you are excited about their work, if the PI is excited about the work then you will be successful.
    If you hate getting up in the morning to go to lab, then you are in the wrong lab.

  • Craig says:

    The PI being involved on a day to day basis was a good thing for me. It allows one to develop a project as they see fit. I do think, however, that many PIs do not appreciate the responsibility of enhancing the lab members careers in addition to their own. In other words, grad students should not languish on ill-planned experiments for years on end and post-docs should only be taken if one anticipates they will be able to help facilitate that post-doc enter into the career of their choice.

  • juniorprof says:

    Roger Nicoll and David Bredt had a wonderful collaboration before David went to Eli Lilly. In his talks, David would always poke fun at Roger because Roger had tremendous intuition but didn't understand a thing about the complicated molecular biology that David is so great at
    I love this example. We have recently been looking over some of Roger Nicoll's old papers on patch clamp experiments in frog sensory neurons (70s and early 80s). It is quite remarkable to look at the progression of ideas coming from his early work (mostly GABA and primary afferent depolarization) into the beginning of his collaboration with Bredt and their discoveries on AMPA receptors. I will have to borrow your anecdote as we discuss Roger Nicoll's early work with our trainees. Thanks for mentioning it!

  • Becca says:

    The system is broken. If not because it doesn't work for so many, than because it wastes so much time for so many. Bad mentorship is a leading cause. To my mind, that is what MsPhD's post is really about- not about PIs doing benchwork (I'm not saying you are attacking a Strawman, PP, just that although it's a good jumping off place for a soapbox piece for you, the out of context quotes may mislead a reader or two).
    *However*, I also recognize the importance of PIs. It is incredibly time consuming and difficult to mentor (particularly for a pairing of personalities that isn't just dead right) and that is made more difficult by many aspects of the system (it's hard to help a green grad student learn the right end of the pippetor if you're always writing grants).
    I don't think PIs need to be at the bench, though it's good for morale if they occasioinally come out and do a little something. That said, it is incredibly necessary PIs become effective delegators if they are to be effective leaders (probably true in any management position). You must have someone teach techniques in a lab, particularly if you notice someone coming into the lab has sub-par performance on a key method for their project. One can learn an enormous amount by carefully watching someone with good hands- and being carefully watched when learning new techniques (particularly when the watcher is supportive of beginners and mistakes). If your PI can't serve that role, help-you-God if there isn't a senior grad student, post-doc, turbo-tech, lab member down the hall, or *somebody* for you.
    Sadly, PP, there are also a lot of PIs who are highly creative, excellent methodologically, perfectly adaquete mentors on meta-skills (like getting their lab members to meet the right people at conferences) and truly lousy at advising for a significant fraction (above a third) of their trainees.
    Maybe they aren't familiar with the sexy new techniques (it *is* hard to keep up; MsPhD is dead on with that), maybe they get *bored* when things don't work for too long, maybe they have favorites of true mini-superstars in the lab and the others languish... none of these imply a totally incompetent mentor (some of these people will be very good for a few of their trainees). These descriptions just describe a human with failings who is not (generally) encouraged by the system to learn exceptional management skills to help overcome these obstacles.
    Benchwork is *hard*. Management is *hard*. It behooves us all to remember both of those things for all interactions between PIs and trainees.

  • CSmonkey says:

    I had the opportunity to work with two mentors in grad school. They were both supportive, so neither of them is a good example of what MsPhD is having to put up with, but they had very different styles.
    Mentor A (MA) was very hands-on in terms of doing the work. If I was stuck at some point, MA was likely to take over. MA also had a tendency to replicate everything I did. In terms of the work we did, we were pretty much at the same level in terms of ideas and work, though MA may have done more work than I did.
    Mentor B (MB) was very involved and available for consultation whenever I needed help too, but if I was stuck, MB would give me some vague, big picture notion (MB was from a different discipline), and I would go back to the drawing board with it. The work I did with MB was more mine, but MB may have done more big picture thinking.
    MB is way more successful than MA, and in terms of what they contributed to my training, it is safe to say that I learned more from MA than from MB about how to be productive, and even about how to think creatively. As a PI, I struggle with my tendency to be like MA. It comes from an underlying feeling that I shouldn't have anyone do work I'm not willing to do myself, but I'd be fooling myself if I thought my trainees are getting much out of that tendency.

  • PhysioProf, if you think that
    Sitting at the bench or having good hands has nothing to do with being a good PI.
    then what's the right way for institutions to evaluate and hire new PIs? You've said that your level of science more or less demands that a new hire have a C/N/S paper plus a couple more decent pubs. Doesn't that select more for people with good hands, rather than PI-ship skills?
    I agree with your basic point that vision and leadership and mentorship are more important qualities for PIs to have, but I don't know how to reconcile that with our hiring system.

  • Alexey says:

    I think many of postdocs and grad students would agreed that very good MENTOR is extremely rare right now in academia. Unfortunately this situation become worse with years. Even at the begin of career when PI is still young and more creative, academic system make him a "whore" of system. That's mean many young PI's become less scientific and creative and don't care much about people in the lab (they are robots - today 10 postdocs left, tomorrow will get 10 new), but they became "heartless administrators", who thinks only about grants, money and benefits - "whores of the system".
    So, how postdocs or PhD candidates should dealing with it? If for 100 of them we have only 1 really good MENTOR?

  • DrugMonkey says:

    becca: The system is broken. If not because it doesn't work for so many, than because it wastes so much time for so many. Bad mentorship is a leading cause. To my mind, that is what MsPhD's post is really about
    Herein lies the rub. It seems as though you are saying (and what YFS is perpetually saying) that the system is "broken" because you are not (yet) a PI or you see very dismal prospects of becoming a PI. And that this is true for "many" postdocs.
    It is a very interesting question of labor politics, vis a vis the pyramid structure we've discussed before. It is an interesting empirical question as to what proportion of postdocs "should" (in a UniversalTruth way) become PIs. Nevertheless, the supply of jobs (and grant $$) versus the supply of postdocs has absolutely nothing to do with individual variation in mentor quality. So there, you are wrong.
    Now if you want to ask on a individual level does the postdoc with bad mentoring stand at disadvantage? Sure. But again, this does not explain the broad trends of a system that is "broken" as you'd have it.
    Dr. J @#8, it takes more than just great hands to get a CNS paper, no? And while I've not been on any hiring committees, everything I know about the hiring process leaves plenty of opportunity to differentiate the candidate who "merely" has good bench skills from that candidate who has a great scientific vision. I think you may be arguing that a scientist with great vision but no CNS paper may be better than some of those with CNS papers. I'd suggest that for any Journal/IF type gating criteria, there will always be those candidates with vision and those with "merely" great bench skill.

  • drdrA says:

    I had the same thought as Dr.J.- If good 'hands' (which are required for good papers, which are required for getting a PI position) are not related to being a great PI later- then are we selecting PIs from the wrong bunch...
    Why should we expect to get a bunch of good leaders, mentors and teachers this way- when we are not selecting from a bunch of good leaders, mentors and teachers??

  • PhysioProf says:

    And while I've not been on any hiring committees, everything I know about the hiring process leaves plenty of opportunity to differentiate the candidate who "merely" has good bench skills from that candidate who has a great scientific vision.

    Having served on a number of hiring committees, I can tell you that this is exactly what we try to do. It is, of course, very difficult. Some candidates with outstanding CVs show up for interviews and within 30 seconds of talking to them, you just know they aren't cut out for the job we have to offer. (Not to say they couldn't be good PIs, but likely not in our environment.)
    Proven ability to lead a C/N/S-level project as evidenced by first-authorship of such a paper is a heuristic for weeding through huge numbers of applications to identify those who will be able to do so in the future. It is an imperfect process to be sure.

  • DrugMonkey says:

    I think many of postdocs and grad students would agreed that very good MENTOR is extremely rare right now in academia.
    What is your definition of a very good mentor? Sometimes I feel like the disgruntled postdoc position is that they insist on a standard of perfection that cannot possible exist. Or if it does, only exceptionally rarely. And refuse to understand that of all the mentoring talents, any given PI is only likely to express some subset.
    Do you really imagine some scenario whereby the laboratory is effortlessly and extensively funded (so you can work on whatever the hell strikes your fancy), operates frequently at the CNS publication level (so you'll get your high impact first author pubs without authorship disputes), is free of scoopage competition (so you can present your data freely at meetings), has a PI who magically roams the international stage and yet still has time to sit down and work at the bench on all the latest techniques...etc?

  • gio says:

    In fact, if anything, there may be some negative correlation. Those PIs who are outstanding experimentalists could be highly effective post-docs without even beginning to develop some of the skills required to be a PI: leading a group, troubleshooting experiments you didn't perform yourself, being creative without being at the bench, etc.

    I found this so interesting. I will soon be a grad student, and I will not hide the fact I would love to become a PI one day, but it always struck me that bench work, which is what most PIs will expect a trainee to do, really does not have much to do with learning how to 1) train people to think critically and interpret manuscripts, 2) train people to effectively write manuscripts and grants 3) offer career guidance 4) teach 5) effectively plan out your research and 6) being basically a good manager.
    Moreover, I feel that students are discouraged from learning the "soft skills" which are still very important to be successful in science, while they are expected to have significant experience at the bench. But in the end, anybody (really, let's be honest) can do some benchwork after working in a lab for a few months, but not anybody might be able to (or willing to) spend time thinking of new research avenues, experiments, read literature and so on.
    Not only the system trains more students than there are really the funds to then support as scientists, but there is also no incentive to diversify your skills and create a proper portfolio!
    Which kind of links to the discussion going on over at Sandra's place.
    I think the truth of the matter is that, as a student/trainee or even postdoc, you need to take responsibility in increasing (as well as widening) your skill set, and probably also for encouraging your university or department (read, advocating) to provide better career guidance for science workers.
    I also think (kind of replying to PhysioProf) that it would be great if experience other than strictly science-related one (say, working for industry in some managing capacity, or experience showing good entrepreneurial skills) could be considered while hiring. Because, as I said before, many tasks a PI will need to be able to complete really have nothing to do with spending an awful lot of time at the bench (which is usually how you become a first author).
    But I am guessing that the funding situation, as well as the fact that universities are unwilling to inject any cash of their own in paying for the researchers (if not the research), make this basically impossible.

  • Liam says:

    Matt posted: If you are excited about the work, if the people around you are excited about their work, if the PI is excited about the work then you will be successful.
    I wholeheartedly agree with this, a positive outlook is critical. But I would place more emphasis on "you" rather than the other people. In my experience, much of research science can be negative in so many ways ("failed" experiments, "rejected" papers, "lost" jobs) that one can easy fall into a spiral of depression. It becomes easy to direct your negative feelings to those around you, including mentors and the real (or maybe perceived) lack of guidance that has contributed to the situation.
    While at times the mentoring I have received has been less than optimal, I have tried to make the best of it and learn something from the process. After all, I am the other half of the equation. Is there something I could have done differently? I have also learned to seek advice from those other than my primary mentor when the need arises.

  • Becca says:

    DM- you've got me all wrong. The system isn't broken because I'm not a PI, or even because it will be difficult for me to become one. The system is broken because I'm extremly smart, highly motivated, and I've almost gotten kicked out of grad school a number of times (I'm on mentor #3).
    It's because I'm immature, I hope (luckily, I'm also young, so I can say that without thinking it implies much about my long term potential or intrinsic character).
    I believe I have a problem with authority, and a fine analytic mind. Science is as good a place as most for that particular combination.
    Ultimately, I'm not of a malicious persuasion- I don't want to make authority look bad just for the sake of doing so, I just believe respect should be earned. I feel like science *should* be a field where I can question away to my heart's content. But the skills I came to grad school with weren't sufficient to let me do that. I've gotten a crash course in communication skills since then, and I'm doing better.
    I identify a lot with what MsPhD writes about on her blog. If you are a hard-core idealist, and believe in honesty, truth-in-data, and egalitarianism... If you can't help but be awestruck at the beauty and wonder of the natural world... then science *sounds* like a great place to work. For some, it is the best job on the planet. But for some of us, it will chew you up and spit you out if you can't put your ideals on the shelf sometimes.
    Almost all the professors I've seen *want* to be great mentors. I believe in their capacity. Yet I'm seeing them fall short. I'm still trying to understand why.
    Most of the grad students I've seen *want* to be great grad students. I believe in their capacity. Yet I'm seeing them fall short. I'm also seeing them get kicked out. I'm seeing this happen disproportionately (as in 11/12) to women. I'm seeing this happen disproprotionately (as in 10/12) to international students. I'm still trying to understand why.
    I realize my program is only one among many- I'm happy to grant that this is not necessarily typical. Yet the most optimistic numbers I've heard suggest only 1/2 of students who start a PhD will finish one. That may be the best that's possible- I don't know. I do know that (as scientists and academics) we are incredibly smart people. I think we can do better- but only if there isn't a totally excessive clinging to the status quo because we are so afraid of job insecurity.
    Yes, DM, the system really is broken.

  • DrugMonkey says:

    It seems as though you think that everyone who manages to get into a graduate training program deserves to be successful and to end up with a PI slot in the end. The fact that this is not the case is your evidence that the system is "broken". Is this not what you are getting at?
    Consider the proposition that the system in fact works (if not perfectly) and that it is a valid screening process to enroll more individuals in level A than can move on to level B. Can you imagine if graduate school admission was trying to do a one-to-one selection for eventual PIs?
    The graduate programs that I've been around have considerably less than the attrition rate that you cite, so frankly that seems a bit weird. One might almost see it as a kindness, though, to drop people out of the pipeline in the first couple of years instead of waiting until 10 years have been "wasted" in pursuit of a PI slot that will never be gained.

  • juniorprof says:

    I've got say, Becca, that I agree with DM, what exactly is this ideal that you are looking for and it is economically feasible in a non-utopian society? Moreover, like DM, I have never heard of attrition rates that even come close to approaching the numbers that you cite. If they are indeed true for your institution I would suggest that there is something seriously wrong with the program.

  • Q says:

    Such high attrition seems more common in "service" departments--in chemistry/physics/molecular biology where they need warm bodies to TA the intro lab classes. When I joined my PhD program, the attrition rate overall was ~50%, 90% for women (apparently they all quit to have children, when I asked). My program needed to have 80-90 TAs per quarter/semester, so TAing the first year was mandatory, even if you had your own money. A few of my incoming colleagues were obviously smart but obviously not cut out for research, so they did their year and quit.
    "Planned attrition" like this is not necessarily an indication of bad mentoring. It is indicative of the problems of large state universities that don't want to hire staff/senior undergrads to teach the intro lab classes, though. The HUGE difference in outcomes between men and women is a mentor problem, and something that should have been a red flag for me. I am stubborn, though.
    I agree with DM and juniorprof about grad school vs. PI positions. There is not enough mentoring in all the world to let everyone who wants an academic PI position have one.
    When I finished my PhD, I had no intention of staying in academia, but I still loved research science. I will return to academia with a TT position this year after a postdoc and then staff position in a government lab, where I found my self some great mentors and role models. Taking the alternative route was a great preparation for me to learn some skills that will come in handy as I set up my own lab. I greatly enhanced my publication record, learned how to set up collaborations with equals and learned how to compete for and run a research program.
    In a pyramid-like system, it is important to learn the skills that will help you differentiate yourself from the masses. You can learn these things outside academia as well. And you CAN move back and forth if you keep publishing and attending conferences to keep your name and work out there.

  • bayman says:

    PhysioProf,
    I think it's great you're emphasizing the need to develop and use the non-science skills required to be an effective scientific leader. But I think you're getting a bit carried away. At the end of the day it's about the science - science that comes out of the lab. You young PIs still have the benefit of being fresh off your post-docs and can escape the drawbacks of disconnecting yourselves from the lab for a while as you learn your politic ropes. But give it 15-20 years or so, and if you're taking your own advice you might find it comes back to bite you in the ass - when people in your lab are working with techniques you'e never tried yourself, when you can no longer judge how funds and lab space can be best allocated to enable your lab minions, when your students stop listening to you because you're so out of touch with lab reality and culture, when you lose all concept of experimental timelines and feasability - you may find yourself wishing you had continued to put in a bit of time at the bench over the years.
    But hey, by that time it probably won't matter - you'll have enough funding to hire 46 post-docs, throw them in the ring and let the survivors take care of the science for you.

  • Becca says:

    DM- That is not what I'm getting at *at all*.
    I am not worried that my (former) fellow students will not be PIs.
    I am worried that they are brillant people who enjoyed science... and who have utterly dropped off the map.
    I think that most of these individuals *deserved* to finish their degrees- or at least deserved to be treated better in their attempt. I'm reasonably confident that most of them (with good mentoring) could have had very successful runs as students, at least.
    I can see your point, that it might be a blessing-in-disguise to get out now if they had a very poor chance of being a PI, EXCEPT being a PI isn't even a common goal for the students I know. As we all know- BEING A PI IN ACADEMIA ISN'T THE ONLY THING ONE MIGHT WANT TO DO WITH A PHD.
    Also, the numbers I gave were not attrition rates- they were based on a count of all the students who told us they had been *kicked out of their lab by their mentor*, within the span of about a year. At our institution, there is no TA-ing; thus, no mentor = no funding. Many of those folks were, de facto, kicked out of grad school.
    As an aside, yes, my institution is F'ed up.
    Anyway, I wasn't asking 'why can't we all be PIs? but 'why can't all PIs be competent mentors'?.
    *Problems with mentors result in students not getting PhDs.
    *Minorities may be at an increased risk of getting kicked out of their labs after problems with their mentors (at least at my institution).
    If this strikes you as a *perfectly* functioning system, than *you* are so broken I don't know what else to say!
    What ideal am I looking for?
    GIVEN the current funding climate, I'd like to see more PIs to hire one Research Associate (as a permanant position, with benefits) to be at the bench for hands-on mentoring, and a single grad student... rather than "trying on" three grad students only to force two of them out when they "have poor bench skills".
    The ideal I am looking for, is for institutions to take a long, hard look at their PI's grant support... assume it will decrease by 5% a year... and only then figure out how many grad students they should admit.
    (We have a real problem with students finding labs at all here, and I realize most of that is hard economic times... but part is also poor planning and total lack of coordination on the behalf of graduate programs and departments. I think when more professors could take on students, the system was more forgiving of poor mentoring- you *could* go someplace else. That's getting tougher.)
    The ideal I am looking for, is for institutions, when determining how many students they should accept, to take a long, hard look at the people they have tenured. Do these people even want students? Can they afford them (in terms of time as well as grant $)?
    The ideal I am looking for, is for tenure boards to consider whether it *matters* if someone has grant dollars, if (over the long haul) they cannot keep anyone in their lab because they are such PITAs to work for.
    I am looking for grad students to be told 'we let you in, we believe you can do it, we will try to help however we can'- as opposed to treated as some kind of migrant workers ('you are a dispensible cog in the machine' [note that both attitudes have some truth, it is a question of how the message you send alters the behavior of the people you're trying to manage])
    The ideal I am looking for, is for all these talented people who are not working efficiently (both on the mentor and the trainee side) to start socializing each other differently...
    *Taking the necessary time spent to develop competency in soft skills is every bit as important as time at the bench.
    *For a mentor, it is as (or more) important to be able to
    advance your trainee's careers than it is to simply squeeze out another C/N/S paper. Not that tenure boards treat them as equally important- yet.
    (note that I do not mean to imply good mentoring and good publishing to be mutually exclusive- far from it! However, I firmly believe a different emphasis than is currently par for the course would be much healthier for science- and scientists- in the long run).
    *gets off soapbox*
    There now. We may disagree about some of the details, but I suspect you'd like to see some of that too.
    But for heaven's sake, don't assume everyone who sees problems with the system is worried only about their own piddly career prospects- or has a sense of entitlement about what the system owes them. Some of us are just the odd flavor of pessimistic seeming idealist who sees problems everywhere... but who *think we can do better*.

  • DrugMonkey says:

    Becca, I'm merely trying to understand what you are on about. That's all. At this point it seems clear to me that your comments are from the perspective of a department with an attrition rate for grad students that is really foreign to me. When I was a grad student, our dept had 10% or fewer fail to attain the Ph.D. and talking with my friends in other departments at the time, this was a pretty good ballpark for many programs. Not 50% by a long margin. The washouts made sense, for the most part. From my perspective, of course. There were a few that didn't work out with a given PI but in most cases they found another person to work with and things thereafter went well.
    If a department has a continued practice of enrolling twice as many grad students then they think will work out (earn the PhD) then yes, I'd say there's a problem of collective mentoring vis a vis the way they conceptualize graduate education.
    If the booting-from-grad-school falls disproportionately on the minorities, women or any other suspect class, yes indeedy there is a problem. We are all aware of these systematic problems around here. I still don't think this justifies an indictment of "bad mentorship" as a "leading cause" of the system being broken. Your brush is too broad.
    Sorry, I don't recollect if you are in an NIH-funded environment? The reason I ask is that if so, this raises the question of a recent phenomenon tied to NIH grant woes. Not to say that departments shouldn't have exercised some foresight. But when funding starts getting tight, PI behavior can start looking really harsh. Problem is, if the money isn't there, people have to be dumped. That is just plain reality. And sad to say, when tough decisions have to be made, the PI is going to go with their best bet for productivity. Not potential productivity, not someone who is "immature but improving" but someone who is producing well now.
    For a mentor, it is as (or more) important to be able to
    advance your trainee's careers than it is to simply squeeze out another C/N/S paper. Not that tenure boards treat them as equally important- yet.

    Huh? You are kidding right? You really expect some junior PI to prioritize graduate student training over getting tenure! c'mon. There's a happy medium here where mutual interests are satisfied but let's not get carried away with ourselves.

  • CStetrapod says:

    Becca: Careful not to confuse problems with your own institution and problems with the system. I'm in CS at a top-40 research institution, and I've seen plenty of all of the things you want to see. PIs here recognize that taking care of their students is going to get them more grant money in the long term. Your institution sounds extremely dysfunctional---far more so than the average institution.
    I liked CSmonkey's post above. My advisor is an MB-type, and it's worked out great for me. I need the freedom; I wouldn't have learned as much with someone looking over my shoulder all the time. I'm going on to become an advisor myself, and I'm worried that I'll either be too much of the MA-type, being too hands-on and too willing to take over, or that I'll decide that since an MB-type worked for me, it must work for everyone. For most students, the ideal is probably somewhere in between.

  • bayman says:

    You really expect some junior PI to prioritize graduate student training over getting tenure! c'mon.
    As I read it, I think Becca was suggesting that the system needs to be fixed so that these two goals are not mutually exclusive, not that PIs should neglect their own survival. Of course we also have to work within the system as it currently exists, but I think we'd all agree there is a serious problem with an academic system that leads scientists with no alternative but to conclude that doing science and mentoring trainees are secondary priorities.

  • Becca says:

    @DM- I am indeed in an NIH funding dominated environment (academic medical center, Big 10 school). My feeling on the situation is that recent NIH funding woes have exercerbated an already sad situation.
    I think that your = or less than 10% attrition is actually unusual. Keeping in mind I have said *nothing* about the attrition rate at my university itself (since nobody seems to have that data)- I based the 50% on something I'd read. A quick googling sugests that it might have been this:
    (from http://books.nap.edu/openbook.php?record_id=5195&page=1)
    "Of the studies currently available, some institutions place graduate attrition at 50 percent for selected fields in the sciences and humanities; others have documented attrition at levels well over 65 percent for some programs. Some attrition will always occur as students progress through demanding research degree programs. Nevertheless, the rates reported by these institutions are considered "high" compared to estimates provided by faculty and deans in 1960 when they placed attrition at 20 to 40 percent.". Of course, the problems with lumping Science and Humanities are huge.
    Indeed, http://www.insidehighered.com/news/2007/07/17/phd suggests that life science PhD completion (at year 10) is as high as 64% (though it's still under 50% at the 6 year mark).
    I'd love to see some really good numbers on this- especially if they are encouraging enough to support the idea that 10% might actually be typical! But I am slightly skeptical as well.
    I'm not saying "bad mentorship" is a "leading cause of the system being broken". I'm saying "bad mentorship is one cause of PhD students giving up". I do think it is a good problem to target- both because it's effects can be egregious (e.g., to the degree minorities are at greater risk) and because I'm optimistic about PIs wanting to be good mentors.
    @bayman- BINGO!
    Yeah, I grok why the aiming-for-tenure crowd will prioritize another C/N/S paper over training students. But once you do have tenure, can you (or should you) continue to justify the same balance on priorities? That's the part that's harder for me to feel empathy toward.

  • acmegirl says:

    Personally, I think Becca is suffering from the all too common disconnect between the idealistic vision of grad school and acedemia in general and the reality of living in the adult world. Nobody DESERVES to graduate simply because they were admitted into graduate school. It's a long, hard road, and you don't get that hood if you don't walk to the end. Period.
    If your PI is a bad mentor, find another mentor. Find two or three, in fact, because no one mentor is ever going to fulfill all your needs. If you find yourself at an institution that does not as a whole provide good mentorship, go to another institution.
    I spent a lot of time and effort researching schools to make sure I would be in the environment that was right for me. Then, once I got into grad school, I spent a lot more time and effort looking for a mentor who was right for me. I could have joined a lab headed by a big name PI who has his fingers in every pie in my field. Instead, I chose to join the lab of a junior PI who showed that he respected me as a person and who was comfortable with my lifestyle (I have two children). I am often very frustrated when I have to listen to classmates who complain about their PI's, only to find out that either they had heard the stories about that person and thought that it wouldn't happen to them, or, they hadn't even bothered to ask around or do any research on the person foe whom they were going to have to work for several year. That is a stupid, careless thing to do. As they say, you have to lie in the bed you make.
    Oh, yes, and when is comes to respect goes both ways - you cannot demand anyone's respect without bringing something to the table. Neither can you expect someone else to earn your respect without working hard to earn theirs.
    I agree that there are some ineffectual, even damaging mentors out there, but I also think that grad students, as the adults they are, need to take some responsibility for their own education.
    Finally, and keeping with the theme of way too long comments, I'd like to bring it back to what PP was ACTUALLY writing about here, which is the idea that in order to be a good mentor or even helpful to a student/postdoc a PI must be fantastic at benchwork. I agree wholeheartedly with PP that the correlation either does not exist or is negative. I have had both the humbling experience of having my PI point out something incredibly obvious that was screwing up my experiments and the exhilarating experience of having an idea that my PI pooh-poohed actually produce interesting results. I think grad students go through a sort of adolescence - where their PI is like the parent and they have to prove that they can function separately from them. Just like how adolescent children like to say their parents don't know anything, so to do grad students like to bitch about their PI who could never understand all the details of their experiments.

  • I've been out of this discussion for a few days so this may seem to be a non sequitur: while the system may be "broken" for every postdoc who assumes they will be able to earn a tenure-track academic faculty position, why is there so little discussion among postdocs of the other kinds of really satisfying applications of one's PhD training. Somewhere (I wish I could find it now) I read that more than 50% of life science PhDs now pursue careers unlike those of their academic mentors - hence, "alternative" careers are now in the majority.
    I'm happy to help any of my trainees, friends, or readers with their career development in the face of the "postdoc wall," but part of that includes an expectation that the recipient of that advice have a somewhat open mind about research/non-research settings, industry vs academic, and geography. I've had to do some very different things in my career than I would have in my twenties because of a convergence of family and personal issues beyond my control.
    Part of being an adult, scientist or not, is realizing that your career and life will not always go as you have dreamed or planned. One must take personal responsibility for recognizing when one is fed up and acting upon it - regardless of whether the fault lies with the PI, trainee, or both - and try to make the current situation work better, pursue new mentors as acmegirl notes, or do something related or unrelated to one's training.
    Life is way too short to be chronically unhappy - but you make your own luck and there are concessions along the way that need to be made that, sadly, may not be aligned with one's twenties/early-thirties view of oneself but are necessary for moving on with one's life. No experience is ever wasted.

  • BugDoc says:

    Although Becca's experience may be somewhat atypical, I think it would be of great benefit for departments to offer mentoring/advising guidelines, and include some consideration of mentoring as part of the tenure package. After all, teaching evaluations from classes are part of our tenure packages, but there seems to be no evaluation as far as I can tell for a major part of our work, which is teaching and mentoring graduate students.
    Clearly, productivity in research and successful grant funding will always be principal considerations in the tenure decision. However, it wouldn't hurt to include at least some evaluation on advising. The main barrier to this process, as I see it, is that it may be difficult to provide fair evaluations. Any ideas from all of you about the best way to evaluate a PI as a mentor?

  • bill says:

    Becca, if being a PI is not a common goal in your grad school cohort, what is? What other careers/options are the people around you aiming for?

  • NeuroStudent says:

    This may be unique to my institution, but I thought that it was common: For tenure here a faculty member has to graduate 1 or more students...for example, recently a professor here got tenure the day her first student defended (last week) and it had been the only thing keeping her from getting tenure earlier (this professor has 2 RO1s and publishes steadily). So, if you judge the quality of mentoring by whether or not a PI can graduate a student then mentoring is kind of factored in (of course, there are some mentoring independent factors, but at least it's something). Therefore, it's hurts pre-tenure PIs to hold students back in anyway, because without graduating a student the PI will never get tenure.

  • whiny says:

    Generally, I like PI's. I only ever had a problem with one, and I consider him an outlier. If you don't like working with your PI, get out and do something about it. No sense in bitching about tenured folks. They're the ones who actually got tenure, after all. Work on yourselves instead.

  • whimple says:

    Why should we expect to get a bunch of good leaders, mentors and teachers this way- when we are not selecting from a bunch of good leaders, mentors and teachers??
    We don't expect that, and we don't care about that either. What we want is:
    1) people that can make us rich (with indirect dollars)
    2) people that can make us famous (with high impact publications)
    ... nothing really here
    ... nothing really here
    ... nothing really here
    28) all that other stuff: teaching, mentoring, leadership, service, training the next generation, etc.
    People that don't think this is the take home message in academia are wrong. People that tell you otherwise are lying to you.
    Now, that being said, can you (as a trainee) make the system work for you? Sure. Pick mentors carefully. Keep your eyes open. Bail when you're in an irretrievably bad situation. Maintain a positive attitude. Try not to be part of the problem. UNDERSTAND HOW IT REALLY WORKS.

  • Ping!
    (in honor of Dr. Free-Ride's tenure....)

  • it's all one in the same says:

    regarding whimple's #28 priority: certainly as a PI, one would want their trainees to be upstanding members of the scientific community. if they suck, it's kind of a reflection on the PI. it's in your best interest to have the best people, since they will be the ones who generate the data for the all-important grants and papers... but there's no need for the PI to hold the trainees' hands every step of the way. that is not exactly what one means by "training."

  • Beaker says:

    Let's dispense with the myth that you can become a PI without having good hands in the lab. We've all heard the stories of successful senior PIs who couldn't pipet correctly, but they are not true. Building from the first comment, here are three examples:
    1. David Bredt, who purified and cloned nitric oxide synthase as a grad student while working with Solomon Snyder. Those early Bredt and Snyder papers have been cited thousands of times. They were not working in some obscure field, it was highly competitive. Who got those experiments to work and scooped the field? Bredt.
    2. Rob Malenka. Go to Medline, search Malneka RC, and go back to those early papers with Nicoll. Malenka obviously went through an intense experimental period while sitting at his rig. He got those experiments to work.
    3. Sir Michael Berridge (guru of calcium signaling). I've heard Berridge criticized for mainly writing reviews and speculating a lot rather than doing experiments. But those early papers he did in the 1970s on the blowfly salivary gland are beautiful in their elegance and explanatory power.
    Conclusion: good hands in the lab are necessary not sufficient to have a chance at making the jump to successful PI. Having said that, none of the experiments described above were technical tour de forces. They are great experiments because they were designed to successfully address important questions: no more, no less. Have a good senior mentor guiding the thinking (Snyder, Nicoll) helps a lot too.

  • PhysioProf says:

    What we want is:
    1) people that can make us rich (with indirect dollars)
    2) people that can make us famous (with high impact publications)

    PIs with half a fucking brain realize very quickly that being a good mentor goes a very long way towards achieving (1) and (2).

    Let's dispense with the myth that you can become a PI without having good hands in the lab.

    It's not a myth. It is reality. The examples you gave are in fields where "hands" are very important for getting experiments to work, and where the experimental paradigms involve a lone experimenter at a rig or in a cold room: physiology and biochemistry. There are other fields where this is not nearly as important, such as genetics, molecular biology, and (more recently) functional genomics.

  • Beaker says:

    PP, we agree that most successful PIs became successful partly by getting the hell away from the pipet rack once they've landed their jobs. As you have emphasized in previous posts, a PI's time is better spent getting funding (chiefly), positioning the direction of the research program, promoting the lab, etc.
    But it is difficult to get through the search committees without that C/S/N high-impact paper. In the vast majority of cases, even today, that requires some talented benchwork. I'd be curious to hear about the exceptions to my generalization--and figure out how they managed to jump over that stage on the way to becoming successful PIs.

  • whimple says:

    PP: PIs with half a fucking brain realize very quickly that being a good mentor goes a very long way towards achieving (1) and (2).
    No, in the general case, this is wrong. Being a good mentor is only an effective strategy if you have no other options. Being a good mentor is ineffective in the long term because you don't have time to wait for your mentoring to pay off. It is much more effective to hire a pile of post-docs who through luck or skill managed to survive and thrive in the grad-school process and can hit the ground running in your lab with minimal need for additional mentoring. If they don't work out, they can be discarded and replaced.
    This is true not only in science, but in any field of endeavor where people have the mobility to leave after they're trained. Look at baseball: all the mentoring happens (if it happens) in the minor leagues. The big-league teams all succeed by buying the players that are already good to go and ditching them if they can't perform.
    ... although that doesn't mean that being a good mentor isn't the morally right thing to do.

  • PhysioProf says:

    Being a good mentor is ineffective in the long term because you don't have time to wait for your mentoring to pay off. It is much more effective to hire a pile of post-docs who through luck or skill managed to survive and thrive in the grad-school process and can hit the ground running in your lab with minimal need for additional mentoring. If they don't work out, they can be discarded and replaced.

    Are you fucking kidding? Where do you come up with this absurd nonsense?
    As a starting junior faculty member working off start-up funds, there are simply not enough resources to "hire a pile of post-docs" who, if they fail, "can be discarded and replaced". And now that even the most talented young investigators are taking 3 years under good circumstances to score the first R01, it is even more the case.
    In a start-up lab, each and every trainee must be effectively mentored to ensure that they produce. Churning through post-docs and grad students may be a feasible strategy for running a successful lab if you have Hughes-level resources, but it is not even close to feasible for a new junior investigator.
    And your baseball analogy is also a fucking joke. There is a huge amount of mentoring that goes on in the major leagues to try to ensure that talented minor leaguers succeed in the bigs. Sheesh.

  • Nick Anthis says:

    Although being a good bench scientist doesn't necessarily correlate with being an effective mentor, it does help with general lab management (in my experience, at least). At the minimum, you want someone who is aware enough of (and interested enough in) what's going on at the bench-level to be able to manage the lab effectively. If the PI is too out of touch with what's going on day-to-day in the lab, he or she at least needs to have a good lab manager or some sort of organizational structure in place. Otherwise it's chaos.

  • PhysioProf says:

    Oh, and in relation to "time to wait", effective mentoring can easily pay off within a 1 year time frame.

  • gio says:

    Nick, for lab management, you hire a technician. At least, that's my experience with the PIs I have worked with...one technician (unless your lab is huge) and you are good to go.
    I would guess that the first year of thesis work would be the crucial one for mentoring purposes. I myself need some shadowing in the beginning, but I take off after a while - which means, I end up doing things on my own anyways. So one does not really need to be breathing on the students' necks for so long that it impacts their performance negatively. It is more a problem of getting a grad student only if you know you can support them in mentoring terms - if you do not have the time, then just do not get one more student.
    It is also important, for a trainee, to understand what kind of person they are, and look for the appropriate lab head: if you need somebody to be hands-on, look for it. If you need to be left completely alone - again, look for the PI you only see once every four months...Both type of people exist, and they can often be as successful. So make sure you look into it. I personally chose a school with rotations, so that I can not only talk to people about their PIs, but also have some time to figure out whether I really want to spend 5-6 years in a certain lab or not.
    Well, given that I am not a de facto grad student yet, we'll see how that goes...and I hope what I said is not going to anger anyone 😉

  • Becca says:

    @acmegirl- I think what you recommend works great for grad students who are not just *adults* but *experienced*. You have to know what kind of people you like to work with. For example, my mother had an awful, heinous boss for her first job. But she didn't know it. She stuck around for a very long time. Years later, somebody asked her "how did you put up with him for that long, nobody lasted under him?!"- her reply "It was my first job! I didn't know all bosses weren't like that". I've had some bad experiences with PIs. None of them were awful people (indeed, I rather like them, and at least one of them rather liked me)- my program has rotations, and I'm not stupid. I think the advisor-student relationship is just naturally complicated (the adolescence analogy is very apt). I've seen a few resources to help out trainees, but I think my institution (and possibly others) could use more to help out mentors. Perhaps we can view the funding squeeze as a good time to look at this?(from the student's side, low funding means fewer labs to rotate in; from the professor's side that means the 'hire the passel o' postdocs and let them figure it out' strategy is much more risky- as PP points out).
    @Neurostudent- yes, my institution formerly had such a policy. I'm not sure why they ditched it- maybe it happened the same time tuition + stipend started costing more than a post-doc (my institution is currently very post-doc heavy).
    @bill- many of my fellow students are aiming at industry. At least one is strongly interested specifically in government.
    One totally awesome thing about my institution is our career day- we get some really interesting folks. I don't think anyone is really *aiming* to be a bioterrorism specialist for the FBI, but we got a lot of people talking to her when she came to give our keynote (her career path has been both facinating and unique).

  • Niewiap says:

    My take on the issue: If it is better for a PI to hire a bunch of postdocs and implement the "sink or swim" strategy, then there is something deeply wrong with the system. In my opinion PIs should be good mentors, because this allows for the most efficient use of available resources. We should not be trying to select for trainees who simply have nerves of steel and thrive in a highly competitive environment, but for those who can most efficiently answer important scientific question. I am not saying that the two traits are necessarily mutually exclusive, but from the point of view of the development of science it is more beneficial to promote a somewhat naive and unseasoned genius rather than a mediocre scientist with amazing survival skills. That's where good mentorship comes in. Survival skills can be learned, the ability to come up with really great ideas cannot.

  • bayman says:

    from the point of view of the development of science..
    Yes! Let's not forget that's what it's all about. Well put Niewiap.

  • DrugMonkey says:

    it is more beneficial to promote a somewhat naive and unseasoned genius rather than a mediocre scientist with amazing survival skills.
    Why? So they can fail to get funding post-startup and drop off the map?
    I don't think anyone is making the equation that you are. I, for one, would advance the point that for any given talent level, the scientist who knows what time it is, and acts in a way that shepherds her career actively, is going to end up better off.
    Abel @ 27 Part of being an adult, scientist or not, is realizing that your career and life will not always go as you have dreamed or planned. One must take personal responsibility for recognizing when one is fed up and acting upon it - regardless of whether the fault lies with the PI, trainee, or both - and try to make the current situation work better,
    Yep. I ended up working in a type of job category that is considerably different from 1) my motivation for going to grad school and 2) my employment vision for all of grad school and into the part of postdoc'ing where I started suspecting I was going to have to leave science altogether.
    Things change. I now do essentially full time that which I've variously thought would be no part, a minor part or at best, the minor cream reward for drudging away at other parts, of the job. Cool, but it has drawbacks and many career components that were front and center at one time had to be put aside.
    I think one of the keys to successfully navigating the system with psyche intact is to broaden that one fixed idea of what the career has to be else all is failure.

  • Niewiap says:

    DM: I agree that you have to be a tough player to stay in the game, but it doesn't mean that you need to be one necessarily when you start up as a grad student or even as a postdoc. Good mentors are there to show a prospective great scientists how the system works and what you need to do in order to make it work to your advantage. If a trainee is intelligent and not totally socially awkward, they will catch on quickly. On the other hand, without good mentoring some really great minds will simply drop out of the system, and their talent will be lost forever to the scientific world. Why the hell was Einstein a patent office worker for most of his early carreer??? Could he have answered even more fundamental questions in physics, had he been discovered early and mentored correctly? How close was he from abandoning science altogether? How many einsteins are there now, who decided that academic life is too complicated for them and perform jobs that are in no way connected to their extraordinary talents? I am simply proposing that by treating trainees like cannon fodder we are selecting for character traits that are not in any way contributing to the progress of knowledge.

  • DrugMonkey says:

    I am simply proposing that by treating trainees like cannon fodder we are selecting for character traits that are not in any way contributing to the progress of knowledge.
    Not all bioscience fields treat the junior folks like cannon fodder. I would suggest that some of the less "hot", perhaps longer established and more traditional fields are in fact hard up for good quality trainees and treat the great ones like gold.
    So the question becomes, why do trainees chose to become cannon fodder? Why flock to the most densely populated fields?

  • bill says:

    If a field is not "hot" it's even harder up for funding than most, and that's saying something -- so I think it's not so much that people "flock" to the hot fields, as that the cooler fields are sparsely populated out of necessity.
    I say this as someone who started out in parasitology, one of the least hot fields I can think of. (Oh, and the cannon fodder mentality was alive and well there, too; being hard up for trainees doesn't seem to make everyone inclined to treat them well.)

  • PhysioProf says:

    I say this as someone who started out in parasitology, one of the least hot fields I can think of.

    Molecular, cellular, and genetic parasitology is a very hot field.

  • acmegirl says:

    Why the hell was Einstein a patent office worker for most of his early carreer??? Could he have answered even more fundamental questions in physics, had he been discovered early and mentored correctly? How close was he from abandoning science altogether?
    Herr Einstein was a patent clerk because he was Jewish, and anti-Semitism was rampant. He also had a family to support, and needed a job. Most people don't realize that working as a patent clerk was ideal for him - assesing the feasablility of inventions often requires an excellent understanding of physics. Yes, the job was beneath him, but also, most people don't realize that he had four groundbreaking papers published in one year (1905) WHILE working at the patent office. It was also during that year that he FINALLY received a PhD, after encountering enormous difficulties finding someone to study under. That's why that year is known as the "Miracle Year". He did not become an associate professor until 1911, again, because not many Universities would hire Jews.
    Many people knew of Einstein's genius and pushed hard for him at every stage of his career. He also had plenty of mentorship. I would say that it is only BECAUSE of how much support he had that he was successful. Oh, and he did have his first wife, also a physicist, proofreading and typing up his papers for him.
    Just sayin'

  • JSinger says:

    In the vast majority of cases, even today, that requires some talented benchwork. I'd be curious to hear about the exceptions to my generalization--and figure out how they managed to jump over that stage on the way to becoming successful PIs.
    I've certainly never encountered these bench-incompetent PI's by whom PhysioProf and MsPhD are surrounded. Everyone I knew who got a research head position in either academia or industry was at least a solid bench scientist, although not necessarily a brilliant one.
    That said, there were people who read papers all day while directing a skilled technician through their experiments. (Which is probably a better predictor of success as a PI than is having great hands yourself.) But even those cases were certainly capable of picking up a Pipetman and generating solid data themselves. They didn't just walk in as grad students and get handed a tech.
    I think that your = or less than 10% attrition is actually unusual. Keeping in mind I have said *nothing* about the attrition rate at my university itself (since nobody seems to have that data)- I based the 50% on something I'd read.
    In other first-jobs-out-of-college, though, what do you imagine is the rate of people deciding that this isn't what they want to do for the rest of their life? I agree that 10% is way low in my experience, but given that we're talking about 22-year-olds with no idea of what a science career is like and no idea on their or anyone else's part whether they'll be good at it, is a 50% changing of mind such a bad thing? It's only "attrition" from the point of view that being an academic PI is the only respectable use of one's life.

  • bill says:

    ...the point of view that being an academic PI is the only respectable use of one's life.
    Which is essentially the motto of research communities everywhere (except perhaps industry, where being an industry PI is presumably the only respectable goal). The idea that pursuing any career other than research is tantamount to failure as a scientist and a person is extremely pervasive.

  • okham says:

    Never throughout my academic experience have I been so bitter, frustrated and angry at the entire system as during my second postdoc. I was seeing my chances of landing a tenure-track faculty appointment dwindling, other people succeeding who, in my own judgment, were not as good as me, and my PI seemingly not caring about my future at all. I could not think of anything else, or talk about anything else. I remember driving my wife nuts (she does not care to talk about that time now), as well as my friends.
    I remember back then thinking that I had it all figured out. I knew exactly what was wrong with academia, how to fix it, and why someone like me (obviously the "best thing after sliced bread") could not find a job.
    More than a decade later, do I now think that all that I was thinking back then was wrong, simply because at some point, mostly out of luck, I had a (positive) reversal of fortunes ? No, I still think that much of that was right -- but there was also much else that I did not know, or appreciate back then. When one gets to be a PI, one realizes what a complex optimization problem that is.
    I try to avoid some of the mistakes that my PIs (and make other ones, of course), but I can also see now what was motivating some of their decision back then (I wish they had been better at explaining it to me).

  • PhysioProf says:

    I've certainly never encountered these bench-incompetent PI's by whom PhysioProf and MsPhD are surrounded. Everyone I knew who got a research head position in either academia or industry was at least a solid bench scientist, although not necessarily a brilliant one.

    I never said I was "surrounded by" PIs who were not good experimentalists. All I said was that they certainly exist.
    Just for the record, I personally was outstanding at the bench, and known for having excellent hands as a grad student and post-doc.
    And Ockham's comment is extremely perceptive.

  • neurolover says:

    This thread is really about "good hands", but we've also started a discussion about the treatment of bench workers (grad students, post-docs, research techs). I'll throw in a problem I have with the system, as currently functioning. I don't think students should be supported by "pseudo jobs" (i.e. TA-ships & RA-ships) where they are funded to perform a service. In an ideal world, those *-ships mesh seamlessly with their graduate degree (as they did for mine). I think grad students should be supported on the Rockefeller model (or what used to be the Rockefeller model) + personal fellowships (like NSF fellowship). NRSA's (NIH's student fellowships) could also play a role, though I feel they're stacked a bit too much towards being awarded to the PI and not the student. Any scheme like this would reduce the number of graduate students, but it would do so recognizing that we shouldn't be using graduate students as low-paid labor with the promise of winning the tournament (i.e. becoming a PI) as part of their compensation.
    Any large scale change like that, though, would be a huge monkey wrench in our current system, though. And, it would weed out many of the students who are currently failing (and some who are succeeding, surprisingly). Our problem with cutting down admittees, rather than seeing them kicked out of labs, is how bad we are at figuring out who will be successful.
    bj

  • Betsy says:

    I agree with DM that the PI shouldn't necessarily have the best hands in the lab. They may have been at one time, but when you're not doing it everyday, you lose your touch. They (hopefully) replace those skills with ones that will make them a better PI. I also agree with some of the comments about having SOMEONE in the lab (ie an RA or lab manager) who can train new students.
    What I always found interesting is that for all the training we have to go through, scientists aren't really well-trained to be PIs. If that's supposed to be the "ultimate goal" of the training, why don't we do a better job of preparing students/postdocs for it? I've always thought there should be some sort of management training--I used to think it should be mandatory for incoming PIs, but maybe it should be part of grad school training. Even if you don't end up as an academic PI, you might end up in industry/government, having to manage a group of people. When you get there, you may realize you have no idea how to do that, because all you've had to manage over the last 10 years is your own research project.
    My postdoc lab was complete chaos, because my PI didn't know how to (or didn't want to) manage the people in her lab. She didn't take the time to hire good RAs, so much of the grunt work fell on the postdocs in the lab, and MUCH tension ensued. Screaming matches were not uncommon, so you can imagine how much "collaboration" went on among members of the lab. I'm convinced that if the lab was run more smoothly on a fundamental level, the science would have been better--but no one ever showed her how to do that.
    Some of this echoes @okham's comment about not realizing what was involved in being a PI until he got there. Why is that? What can we change about the training process so that transition happens more naturally?

  • DrugMonkey says:

    I don't think students should be supported by "pseudo jobs" (i.e. TA-ships & RA-ships) where they are funded to perform a service.
    I will take your point that in some cases the use of graduate students to TA endless class hours is unjustifiably exploitative. My grad program expected 2/3 quarters and everyone picked up a summer TAship 'cause the pay was so dismal. People with fellowships still were required one TA per year- uncompensated, which was clearly bogus.
    Nevertheless, I would probably have little difficulty in saying that TAing (not "you teach the whole damn class" but really assisting) once a year in grad school is not a bad thing. One is supposed to be training to be a professor and getting some real world experience in this job is a GoodThing all around.
    The other thing grad students are supposed to be learning is the job of being a scientist. Not, "to conduct whatever research I fancy on whatever schedule I fancy". IMHO, of course. So I don't see where there is any problem whatsoever with RAships, if you mean under the advisor's grants. I mean, you are being paid to work in your selected research lab...isn't this the goal? To have a career being paid to work in your selected research lab (hopefully with you as PI)?
    Where's the problem with RAships and, really, how is this solved with fellowships? Do you really think that RA versus NRSA is really going to change how your PI behaves? If S/he is a slavedriving micromanager, waving your NRSA around isn't going to help.

  • PhysioProf says:

    Do you really think that RA versus NRSA is really going to change how your PI behaves? If S/he is a slavedriving micromanager, waving your NRSA around isn't going to help.

    For post-docs, there is a big different in how those with their own fellowships versus those paid from PI research grants are treated. This was true for me as a post-doc, is true for the post-docs in my own lab, and is true for all post-docs and PIs I have ever discussed this with.
    If you have your own post-doc fellowship, you get a lot more leeway in exactly what research goals you pursue. If you are paid from a research grant, you pursue the goals of the research grant, as determined by the PI.

  • whimple says:

    If you have your own post-doc fellowship, you get a lot more leeway in exactly what research goals you pursue.
    This may be field specific, because in my postdoctoral experience the source of postdoctoral funding was largely irrelevant. (Of course, funding cash was also easier to come by at that time so fellowship or not wasn't that big of a deal.)

  • JSinger says:

    This may be field specific, because in my postdoctoral experience the source of postdoctoral funding was largely irrelevant.
    Same here. The only obvious difference was that winning the fellowship cost me my medical benefits. Given the trivial amount of my pre-fellowship salary relative to the cost of animals and reagents (the mice, of course, continued to receive medical coverage) I can't imagine that not having to pay it mattered much to the PI deciding what research I could pursue.

  • Enrique says:

    What would you recommend for a postdoc wishing to start a project if his/her own? Is this a good time to delegate wet-lab responsibility (assign some initial work - cloning - to a technician willing to do the work)? Or should he/she work up everything on their own? I am apprehensive to delegate, because the technician almost never takes my advice when problems arise (because I am not "the boss") ... but, then again, I am currently swamped at work and home.

  • DrugMonkey says:

    What would you recommend for a postdoc wishing to start a project if his/her own? Is this a good time to delegate wet-lab responsibility
    In a word, yes. I see it as a continual process of acquiring the new skills that you need. One of these is learning how to employ and manage technical staff for your projects. Everyone's subfields are different so I can't give specifics. The idea as I see it is to identify those tasks which you can train your available help to perform. That help may be undergraduate assistants, rotation students, techs or even grad students. The idea is that you are starting to become responsible for leading no matter what the status or expectations of those types of people that are available to you. (and, let us be frank, in managing in the sense of maximizing an individual's skills, talents, and or limitations no matter what you expect they are supposed to be able to do a priori!)

  • Enrique says:

    In a word, yes ... Everyone's subfields are different so I can't give specifics. I work in a field where one can get away without extensive wet-lab skills. I was going to use this summer to build up my wet-lab skills ... start my own project ... prepare a grant application ... as well as continue to do my current work, plus some new work I am starting for my second phase of postdoc. Then again, I wonder how much I would actually use these wet-lab skills later on when I find a "real" job.

  • Enrique says:

    Also, the wet-lab work I would delegate now is the same type of work I will (likely) be doing in ~2-3 months when I start my second phase of postdoc. Delegate now = results sooner. Delegate later = do the work myself and learn.

  • PhysioProf says:

    Dude, if you got trainees you are authorized to mentor, then start fucking mentoring!

  • Enrique says:

    Dude, if you got trainees you are authorized to mentor, then start fucking mentoring!
    I take it that means get my head out of my ass and delegate ... in PP speak.

  • PhysioProf says:

    There's a lot more to mentoring than "delegating".

  • Enrique says:

    Well, the technician (and graduate student) I work with (and "mentor") aren't really my "employees" to mentor in the first place. Unless you mean I should be taking the bull by the horns, treating them (in my mind) as I am their boss and mentor for the purposes of my training.

  • Beaker says:

    Wow, this thread turned into about four, four threads in one. Hats off to PhysioProf for pushing so many buttons in one post. And it didn't even require the creative use of profanity. Blogafuckingeloquent!

  • juniorprof says:

    Enrique,
    Why don't you forget about who's boss and who employs who and lead the project in such a way that everyone will learn something and the scientific goals will be advanced to everyone's advantage. That, I think, would be effective mentoring.

  • neurolover says:

    "The other thing grad students are supposed to be learning is the job of being a scientist. Not, "to conduct whatever research I fancy on whatever schedule I fancy"."
    Hmh, we may have discovered something fundamentally wrong with *my* training :-). Honestly, I because a scientist so that "I could conduct research I fancy on whatever schedule I fancy." (I do realize that may have been, shall we say, a slightly fanciful vision of the enterprise?)
    But, I still think the problem I have with RA-ships (and TA-ships) is the confound of the job and the training. The job part is where you are providing a service; the training part is where you are learning something. Sometimes both of them can happen together, and sometimes they don't, and sometimes they actively conflict. If RA-ships are a job, a scientist job, they should be a job, not, for example, restricted only to students who are enrolled in a graduate program. Not using students (a captive audience for this labor) would force universities to think about what they're requiring for the job, and what they're doing for the training.
    (And, I feel this way about soft money positions at universities as well -- I think they create incentives for employers to behave badly).

Leave a Reply