Mistakes I've Made as a PI: Mentoring Trainees

Aug 17 2016 Published by under Uncategorized

The current mistake was in how I hired students. It's relevant to current Postdoc Salary issues, to which I'll return at the end of this post.

In the summer, we (my department, other labs in the med school here) take summer students into our labs. Some of these are med students, some other clinical professional students or college students, and there's even the occasional high school student. These folks can be paid in a couple of ways, but the two big ones are on a fellowship (an amount regulated by the university that isn't very good, but its better than volunteering, something to which I object) or as an hourly employee. NIH grants will not pay student fellowships. People paid as part time on an NIH grant need to be an employee, and make at least minimum wage. Fellowships, by the way, work out to less than minimum wage in my lab, where the students are seriously involved in the work.

I make all of this very clear, including different categories (and when possible offer students the choice), before students start. I tell them, no I discuss with them, what working in my lab entails: the hours are long, some of the work is less tedious (data collection), some of it exciting. I tell them the upsides: fellowship students can go to a national meeting, at my expense, if they get an abstract accepted (and all who have submitted one have been accepted), their name goes on the papers to which they contribute. Finally, I give them names of former summer students and tell them to go talk to them. I trust the former people to give accurate information, good and bad.

This summer, I had two students on fellowships. The third person, Jane, was hourly, and on my grant. All students worked long and hard hours, as is the case with large animal work. The first two were wonderful, did well, and will get their names on at least one paper. They are back to their student lives this week, and I will miss them.

These two were acutely aware that Jane was making more than they were. Jane turned into a disaster (which is another glorious post all on its own).  The disaster was not because Jane was in a different employment category. The disaster was Jane not being honest. But even if Jane had not been a disaster, the two categories was the basis of my mistake. Having two categories made for bad feelings, and a number of less than totally smooth incidents in the lab.

It may be possible to have two categories. After all we have postdocs and grad students and technicians of various levels and skills. My mistake was not differentiating between them more cleanly. And having not distinguished between these two categories might have contributed to the disaster, but it also could have happened anyway. I won't know. Life is not a ceteris paribus experiment.

It seems obvious now. People doing the same job, with roughly the same experience, should not be paid significantly different amounts of money. I should have distinguished between the jobs this summer, and there were lots of ways to do that. For example, fellowship people work fewer hours. Fellowship people do more reading, more development of ideas, more presentations in journal club. Fellowship people get  "a project". Hourly people wash bottles, and take on more grunt work. But making this distinction is not necessarily easy, and it's obvious only in hindsight. If one gives fellowship people a "good experience" one can end up spending more time designing, implementing and assessing the experience, and not getting the science done.

Why is this relevant to postdoc salary issues? Well, my med school, wrestling with the postdoc hours/salary problem has decided, will likely decide, to have two categories of post-degree research employment, names to be determined. In one, with salaries below the $47K threshold, there will be set hours and, allegedly, set tasks. Overtime will be limited to what the PI (who is carrying the salary) can afford, but in general discouraged. These people may have to clock in and out to demonstrate hours worked. The other category, over the threshold, will be able to (what a verb) work unlimited hours. There are going to be two categories of job, two categories of postdoc, and ultimately two distinct duties/assignments. Leadership believes that distinguishing between these categories will be possible and produce only minor problems.

I know that there are currently different levels of pay for postdocs. Some differences are due to age and seniority, some due to cheapness of PI, some due to passport-of-postdoc-origin. I am not defending nor condemning those. But the proposed scheme will be different in that two postdocs, maybe even in the same lab, of roughly same experience may end up with distinctly different jobs at distinctly different pay. Leadership often does not perceive the problems that PIs wrestle with in the trenches between benches.

However! Fear Not! There are ways to cope in advance, things that are worthwhile in general being a mentor.  Lay out specific project and duties. One of the best is to develop an IDP for each person (which probably is important to do anyway). Here's one from FASEB. Here's some NIH info on IDP and postdoc success. And the SCIENCE careers page which has lots of stuff on IDPs, including tools for the postdoc to use.

I'm not sure this would work for summer students who are here for 10-12 weeks in different categories. I suspect in the future, I'll solve this problem by having one or the other employment, and avoid the problem. Which, of course, will just free me up to have other problems.


31 responses so far

  • drugmonkey says:

    Really good reminder and insight. Thanks.

    ....and dear sweet baby Jebus please let my administration not decide to do this.

  • Ola says:

    Thankfully our admin just decided all postdocs will earn above the 47k threshold, no exceptions. This is sort of in line with their previous rule that all post-docs will be paid at a level no less than 90% of the prevailing NRSA stipend for years served. They also moved, several years ago, to eliminate the problem of people losing out on benefits when they won a fellowship - they united the pay codes so that post-doc employees vs. fellows are all the same, same benefits, everything.

    If things are different at your institution, the real question is who are the skinflint PIs lobbying the administration for a 2 tier system? Why has the local PDA (if one exists) not been effective at ending such practices?

    • potnia theron says:

      Indeed, some of the objections are from tightwads. No sympathy there. But also no PDA to deal with this.

      One real problem is people with NSF grants. These are usually an order of magnitude smaller than NIH grants. There is almost no wiggle room (no PI salary permitted). If these postdocs are in the mid 30's range, there is just no money to push them over the limit. I think some (temporary) accommodation is defensible here.

  • GM says:

    I am curious how exactly the clocking in and out will be enforced.

    I do experiments and I also do a lot of bioinformatic analysis, in a something like a 30-70 ratio.

    I can imagine ways to clock in and out the experimental work, but who exactly is going to clock in and out the bioinformatic analysis work, and how? I do that at all times of the day, and at all sorts of places, as long as I have an internet connection and have access to the cluster. Are they going to prevent me from accessing the cluster outside the 9-to-5 hours?

    Then there is paper writing, peer review and other activities of the sort that people do on their own laptops. This is absolutely impossible to track. But it is still real work.

    Of course, I would like to think I would be put in the "unlimited-hours" category if something like this was to be enforced more widely.

    But the general point remains -- science is by its very nature ill suited to 9-to-5 work practices.

    • potnia theron says:

      That's part of the point of raising salaries over the threshold. Problems such as these go away when that happens. I am most concerned about PI's who won't (but in some cases, such as NSF funding can't) go over the threshold. Coercion is a real concern.

    • Eli Rabett says:

      Your bioinformetics package should be able to track your hours, if not there is an app.

  • acd says:

    Some undergraduates *need* the extra cash that an hourly appointment provides for tuition, books, etc., and perhaps by delegating "grunt work" to these hourly employees, the underprivileged students are not put on an equal playing field with those who are in a position to accept a fellowship. (Of course this may be a bit different situation for postdocs.)

    • potnia theron says:

      This is the problem with "free" "volunteer" and otherwise underpaid fellowships. But its a problem that exists across all professional fields, including scientific ones. People with more money can buy things for themselves, or their children that people without money cannot.

  • Anon says:

    NSF is allowing re-budgeting, and left a teaser of POTENTIALLY asking for supplement but word in an FAQ is different that actually following through.


  • JohnBorghi says:

    Thanks for this.

    I observed something sort of similar as a PhD student. I was funded primarily through TA lines, as were the majority of labmates, as were the majority of students in the department. The major exception were people who had received fellowships through an underrepresented students program.

    There were bad feelings when two MA students received the fellowship and didn't have to TA, there were more bad feelings when it was revealed that the fellowship was almost 2x as much as the TA stipend, and there were really ugly bad feelings when one of the students (who happened to work in my lab) turned out to be a total disaster.

    • potnia theron says:

      I agree about bad feelings. For grad students, it undermines a sense of collegiality, willingness to help each other and share data.

      I put the onus of making this right and work on the PI.

      • JohnBorghi says:

        I agree- even if it's just a simple "Hey, maybe don't talk so loudly about how much more you're making then everyone else.".

        In defense of my PI, she generally tried to maintain a good lab culture. Her efforts were just undermined by a lot of factors only tangentially under her control- SUNY's nonsensical pay structure for TAs, the department suddenly admitting MA students after the 2008 economy collapse, etc

  • drugmonkey says:

    What are the odds that more scientists want to work odd hours than have to?

    Also- how much resistance comes from a sneaking suspicion that trainees don't actually work 40 h a week. Putting the will and structure to clock actual hours (don't lawyers do this to every 15 min?) might be uncomfortably revealing.

    • GM says:

      I never said I am working 80h a week, just that if you try to force a rigid schedule on me, that will completely destroy my productivity.

      That sort of thing:


      Happens a lot. That is true.

      But I personally work best when I feel my scientific work as inseparable from my life (the worst periods of my life have been where for various reasons there was some separation between those two things). And I know others feel the same way.

      What the clocking in and out will do is reinforce the separation between those, who are actually living the scientific life (if we can still speak of such a thing), and those working for them, who are basically drones.

      Of course, it's already like that in practice in many places, therefore it's possible that making it more explicit might not be a bad idea.

    • I totally agree--when I actually had to work 60+ hours per week to meet a deadline, I didn't have time to wash my clothes, arrange/go to lunch, deal with my snail mail, or any of the other tasks needed for modern life (thankfully, I had ProdigalSpouse to help run my life). All my time was spent working, commuting, sleeping, eating, or on personal hygiene. It was completely unsustainable, and made me realize how unlikely it is that many of the people who claim 60+ hour work weeks actually work that much on a regular basis (face time is another thing...).

      In grad school, I was easily able to schedule my time to avoid working weekends (I don't work with anything alive, which helps with that) and still had time for a social life. My most productive students are the ones who work at work, and do other things on their own time, even though they spend less time in the lab. I suspect that most postdocs could fit their work into 40 hours if they focus and the culture allows it.

      • jmz4 says:

        60 hours a week is less than 9 hours a day, if you're working weekends....just how long do you shower?

        "(I don't work with anything alive, which helps with that)"
        -That is a giant difference. I spend almost 10 hours a week keeping my various critters and parts of critters alive. Of course, I also have the scientific equivalent of hoarding. I have way more strains and lines than I actually need at any given time (but just might next week).

        Also, I could easily fill all the hours in a week if I were working on all my projects simultaneously.

        And again, it boils down to what is "working"? Commenting on blog posts, probably not. Reading papers on the bus? Probably yes, since I'm expected to be knowledgeable in my field. Does a failed Western blot that I'm ashamed to tell my boss about (cause I plugged the wires in wrong) count?

        The people I know that get a lot of stuff done work regular, long hours. They're not the people like me, who are physically in lab a lot of time, but not extremely focused. They come in from 9-8, every day, and maybe take a Sunday off. These are the 2-3 paper a year people.

  • The Other Side says:

    As a manager, how much sense does it make to take away one of the few perks of being a postdoc (flexibility with hours) when postdocs make a $hit salary as is? These are the people you expect to go above and beyond, right?

    "My most productive students are the ones who work at work, and do other things on their own time, even though they spend less time in the lab."

    Good for them. I know plenty of people to whom this does not apply. They happen to be primarily concentrated in Academia -- I think this is no coincidence.

    I wonder how PIs would feel if they had to track their hours, just like the lawyers.

    • potnia theron says:

      It is a useful exercise for anyone, at any level, to spend a week tracking their hours. read the post I put in the earlier comment.

      • The Other Side says:

        I've already read your post. Tracking your hours as an exercise for yourself is a completely different thing than having it be the basis on which you are paid.

        Again, if this is such a great idea, why aren't faculty clamoring to do it?

  • Nat says:

    I love the academic view of labor markets and regulation. Amusing, at least now that I don't have to deal with the nonsense, having moved to industry.

    No one should be surprised that if people are doing substantially the same work, and they are receiving different pay/benefits/perks, bad feelings are likely to ensue. This is simply human nature. So, having 2 tiers of postdocs or summer students is ripe for this sort of problem.

    Relatedly, there's the problem of bad acting PIs abusing the system (trying to pay post-docs at the cheaper level, while expecting an unreasonable amount of work. Or, what about hiring foreign post-docs at the cheaper level, and then expecting long hours of a person who has little knowledge of U.S. regulations, and who is dependent on their PI for their visa status).

    The productivity effects of working >40 hours per week are extremely well studied and documented in probably every line of work studied. You may think you're getting more done, but you probably aren't (even if you are working much more than ~45 hours/week, which is questionable). I don't see any reason to expect that academic work is different than other knowledge work in consulting/medicine/legal industries. But, actually measuring what productivity is in the academic setting is difficult, which makes analyzing this tricky. That perpetuates the myth of the overworking academic, which is seen by some as a prerequisite.

    The Other Side: "As a manager, how much sense does it make to take away one of the few perks of being a postdoc (flexibility with hours) when postdocs make a $hit salary as is? These are the people you expect to go above and beyond, right?"

    Flexible hours and requiring overtime pay for >40 hours worked per week are completely unrelated issues. You could allow someone total flexibility to choose which hours they work, but if they are working >40 hours per week, they should be paid overtime. And if employers don't want to track that, then pay them the salary at which overtime no longer applies.

    GM "What the clocking in and out will do is reinforce the separation between those, who are actually living the scientific life (if we can still speak of such a thing), and those working for them, who are basically drones."

    This notion of a "scientific life" and that academia is concentrated with "special" type of scientists is total hogwash. I spent my whole life in academia and now work in industry- it ain't different.

    Talking about people as "basically drones" is pretty awful. Maybe you didn't intend for it to sound as bad as it does, but oy.

    And as for tracking hours, here's a news flash, people all over industry do it. When your efforts go to different projects, and the cost of those projects has to be tracked, then you track your hours. I do it, my boss and their boss does it, the people who report to me do it.

    • GM says:

      Talking about people as "basically drones" is pretty awful.

      I am describing the way trainees are viewed by many PIs.

      I hate it just as much as you do.

      Regarding tracking hours:

      If actual bench work was all there was to being a postdoc, then it might have been possible to do that (although even then there are complications -- for example, does reading the news online during that 30-minute incubation that is too short to do anything meaningful directly related to work count as "not working"?).

      But it is not.

      You have to analyze data, write papers, think about things, etc. None of which can be tracked in any meaningful way. Does all the thinking I've done while lying in bed, or just sitting somewhere away from the lab, or walking to the store to do the groceries, etc. etc. count as "work"? Probably not. But without that, a long list of projects that were successfully completed would have gotten nowhere.

      The point is that if working in science is to be seen as intellectual work, then it is impossible to clock in and clock out, because nobody can force anyone to stop thinking about their research when the clock is not running. And forcing people to think from 9-to-5 is not exactly a good way to stimulate them to think well. Also note that people's perspective here is being heavily skewed by most of them working in biomedical research and chemistry, where there is a lot of bench work -- there are many fields where thinking and writing is all that people do.

      The implementation of clocking in and clocking out would send a clear message that the work of postdocs is not seen as having any intellectual component to it. Thus my comment about "drones".

      • Nat says:

        I'm not clear, is it many PIs or you who view clocking in and out as being a drone? Cause between the first and last part of this comment, I'm finding the distinction a bit fuzzy. But it's probably not worth harping on. If your PI (or you), can't see the potential human capital in their employees who are on the clock, then whatevs.

        Nonetheless, I fully agree that the typical job description of a post-doc entails all the things you list- I was a post-doc for a while ;). And it's a level of job duties that suggests it should be carried out by a salaried employee. So, make those people salaried employees.

  • The Other Side says:

    @Nat: "And as for tracking hours, here's a news flash, people all over industry do it."

    Yes, and having worked in industry for over a decade before going into Academia, I am aware of how many people "do it." It's bullshit! Forcing postdocs to engage in the same kind of bullshit is no solution.

    Certain jobs are exempt for a reason. Being a postdoc should be one of them. And yes, postdocs should make more, even if it means there are fewer of them.

    @ GM: "The point is that if working in science is to be seen as intellectual work, then it is impossible to clock in and clock out, because nobody can force anyone to stop thinking about their research when the clock is not running."

    Exactly! But again, I say that it is extremely revealing that not a single faculty member I know is advocating that *their* hours should be tracked. In fact, some have a hard time with electronic calendars because they are so loathe to declare to their colleagues when they are or aren't available!

    If you, as faculty, don't think that clocking in and out would make you more efficient, etc., why do you think it would have that impact on your postdoc? You have traded a higher salary for the freedoms of the academy, no? Well your postdoc has done that in spades! Stop pretending you don't know that.

    • Nat says:

      Yeah, I'm in favor of making post-docs exempt from this sort of time reporting requirement- by having their salary increased past the exemption level. I'm not in favor of any special pleading that post-docs are different than other jobs, and so should have a lower level of salary where the exemption kicks in.

    • If postdocs go on the 40 hour clock, it isn't to make them more efficient. It is to grant them the legal protections to which they are entitled. As Nat said, there is lots of data suggesting that productivity drops off pretty steeply as hours climb much above that anyway, and that matches my personal experience. Postdocs can still be flexible working 40 hours. Flexibility and clocking time are separate issues.

      Practically, if there is no money to pay a postdoc $47k (like for NSF funded researchers), there will be no postdocs unless it is possible to pay postdocs less and still meet legal requirements. It is already significantly harder to find a postdoc in my field than it was even 5 years ago (I am not in biomed). I have never had a postdoc in my group because I cannot pay a decent salary to one. This may be a good thing in the end, but as the changes percolate through the system, there is a lot of pain. My understanding is that biomed research is far more dependent on postdocs than my field is, so there will likely be more pain there as well. You may not want to be on the clock, but it is better than not having a postdoc at all, right?

      Tracking hours isn't the Thought Police. No one can stop me from thinking about work while I am in the shower, just like no one can stop me from thinking about what I should make for dinner while I am in the office. I think anyone who likes their job thinks about it "off the clock" so to speak. That isn't unique to academia.

      Getting back to the original post, I definitely think that pay disparities between group members at the same stage is a really bad idea. I don't think I would ever attempt to have 2 sorts of postdocs (if I could even afford to do such a thing), since it is sure to breed resentment. At the same time, I couldn't afford postdocs BEFORE the rule change, so this won't impact my group in any way. I can understand the institutional desire to accommodate NSF-funded researchers, which might lead to the institution allowing 2 kinds of postdoc. But it would be foolish to have both kinds in the same group.

      • The Other Side says:

        "You may not want to be on the clock, but it is better than not having a postdoc at all, right?"

        Not so sure about that....

  • Eli Rabett says:

    Given a choice btw 40 K for 40 hours (that's $20/hr and 2 wks vacation) and 47 K for unlimited time, there are lots of people (or at least a fair number) who might opt for having a real life. Their choice

  • […] have frequently written about problems with trainees and what to do about Problem […]

Leave a Reply