Can we measure a lab member’s productivity?

I need to fire a lab member; this has kept me up many nights for the past six months. This individual is a staff member in my group, and hasn’t met expectations since the beginning. Yet it’s taken me a long time to finally conclude that this person just has to go, and this conclusion comes after trying different management tactics (more in-person meetings, fewer in-person meetings, increasing responsibilities/independence, switching projects, delineating weekly goals, increasing specificity of goals). We have had multiple performance reviews and discussed that this individual isn’t meeting expectations; we both go into and come out of these conversations with good intentions and outlined changes we will both make, but then a few weeks later deadlines will be missed. A couple of key opportunities for presenting results and gaining new skills were not followed through by this lab member. We are simply not a good fit as a supervisor/employee pair.

Unfortunately this is not the first member of my lab I’ve had to terminate. I thank my lucky stars that none of these individuals have been graduate students or postdocs. My graduate and postdoctoral trainees have been wonderful lab members, and I have learned quite a bit from each of them scientifically and as a manager.

A huge part of me wonders if and how I might have better managed my unsuccessful management relationships. And that same part of me wonders if my knee-jerk internal monologue — “If I just changed something about my approach, this person would be successful at his/her job” — is something I should just push aside. For lack of a better way to put it, I wonder “Would any of my advisors think twice about firing these employees?”

What do these three cases have in common? Each case had similar small-scale logistical problems that were part of a larger issue: for example, lack of attention to detail; not taking notes during meetings with me (it is so strange to me to have to tell a lab member “You should be writing this down” but I have had to do that multiple times); missing deadlines and giving no explanation as to why, and giving no indication that the deadline wasn’t going to be met; not applying suggestions I made about moving forward with experiments and instead reverting to previous approaches that were known to not work. The larger issue all of these examples reveal was a lack of “productivity”. What I’ve realized in managing a lab for a few years is that unfortunately productivity is hard to measure or define: like Justice Potter Stewart said “I know it when I see it.” For a trainee who is meeting goals and moving work forward, I don’t mind if they keep strange hours or are sometimes late to meetings or need to cancel on me once in a while. When a lab member isn’t meeting goals over multiple weeks/months, something as simple as that lab member seeming distracted during lab meeting can have me seeing red.

Can we measure productivity?  I have this idea that this is easier in other industries, but it seems quite hard in academics. Number of publications? Accepted abstracts at meetings? Figures for our one-on-one meetings? When I think about how often I am reviewed by my colleagues (every year nominally, but really every few years) and the expectations my institution has for me (“for promotion to associate professor with tenure, the candidate must develop a national reputation in his/her field”), it’s no wonder that I haven’t a clue how to outline concrete short-term steps a lab member can take to be labeled “productive” or “successful”.

After much thinking, the two characteristics my thriving lab members have are a combination of taking direction and showing creative initiative in their research. I think the skill we learn as graduate students, and hone as postdocs, is the intuition for good research problems and paths forward when hitting research obstacles. I strive to gently guide my students through obstacles so that, over time, they’ll learn their own intuition when faced with research challenges. These thriving lab members take my advice when we meet about new experiments to conduct. And they will sometimes go above and beyond my advice by learning a new skill, trying a new approach, or thinking about our data in a way that I didn’t. Every so often they will come into a meeting and say, “I did the things we talked about, and tried this new thing.” And those meetings are usually exciting ones. Another way to think about this creative initiative is pursuing things that feel strange or unexpected in the data we are analyzing or theory we are developing – like Thomas Kuhn suggests we scientists do more than our textbooks would reflect.

What I haven’t figured out is how to guide someone into a productive mode when they’re struggling. I see gaining this skill as a manager as vital, as my research program is transitioning from projects I drive forward to projects driven by my advisees. I have one colleague who swears by sitting down with the struggling lab member and working through the problem together, as long as it takes. I have tried to do this, but find it a struggle to help the trainee move forward while not simply doing the work for them (i.e., I need to think more about “teaching how to fish” in some of these instances). I have tried partnering lab members up, but my lab is still small, so this can sometimes burden a couple of go-to folks in the lab. My colleagues have these struggles too — we often grasp at straws to explain why a student is floundering, and then focus on little things like number of hours worked in the lab versus at home. Do my interviewing methods need to be improved while hiring? How can I predict based on a short series of interviews the working relationship I will have with someone over months and years of time?

I don’t have an answer to the title of this post; while I think delineating expectations for PIs and trainees is a worthwhile process/noble goal but I’ve found it difficult to really get at the heart of what my expectations are for lab members (except “productivity”, ugh). One piece of solace I have is that my graduate advisor, whom I dearly admire, told me that he still can’t predict a graduate student’s trajectory, even after over 30 years as a graduate advisor. This reminds me that the scientific training that takes place in labs is a living thing that morphs with the people involved, not some rote set of rules on stone tablets that make for a “good” or “bad” lab experience.

Are there ways productivity is measured in your research group? How have you handled trainees who aren’t meeting expectations? If you struggled during your training (we all have, really; graduate school is too long to not have some lags in the research process), how did your advisor handle it? How do you wish your advisor had handled it?

8 thoughts on “Can we measure a lab member’s productivity?

  1. What strikes me about your post is that it seems like you are lumping your staff with your advisees (i.e., grad students, post docs) when it comes to your expectations. I imagine your advisees are self-motivated, ambitious, and creative, as indicated by the nature of their career choices. They want to make a name for themselves and impress you, and they do that by being productive. But a lab manager-type–if they don’t intend on going forward in their career by publishing, etc.–has little incentive to do things you suggest, unless they are rewarded with something meaningful to them. –Is your lab manager well-paid? If not, how do you reward him/her for going the extra km / mile? You might want to look at it from that angle (since this person is not the first to have problems under your management = recurrent loss of your time and energy).

    • This is a fair point, although the two staff members in question came to my group intending to work on research for a couple of years before going on to a PhD. In both cases, they wanted to contribute such that they would be authors on papers. So I think their ambitions were in line with advisees’ ambitions. Also, they were well-compensated on an academic scale (full benefits + salaries greater than that of postdocs).

      Your comment reminds me that similar struggles can happen when graduate or postdoctoral advisees at a research university are not seeking research university jobs after completing their training; I know it can be a struggle for advisors to offer professional development for non-academic careers. In my lab members’ cases, all have stated they want university-based professorships, and I do check in about that a couple of times a year to make sure my advising isn’t off the mark.

  2. Let me throw my hat in as the person who’s been on the other end of this situation. In my case, I was a grad student working with a lab that mutually understood I would not be a permanent member, but they had an interesting image processing project they wanted done that I had the experience to tackle better than any of the other lab members. This arrangement dragged on for about a year.

    While i did make progress and learn a lot, many deadlines and expectations were not met. There are many reasons for that, but I do acknowledge that ‘I’ was the one at fault for not meeting the deadlines and expectations of the project – I don’t blame the PI or either of the post-docs I worked with. In my case, there were issues with communication and depression. I was, and still am, immensely interested in what the project is meant to do, but no matter how ‘content’ I felt with the work I did, I never really experienced any joy in that project, or anything else, for that matter. Couple that with the restrictive and closed-loop workflow set by my post-doc, and I felt like I was isolated on a project where the methods were too conflicting with what I felt would’ve been a better way. To clarify, this was a biology lab and all the members were skilled with benchwork, compared to me, who is a computer scientist by training. It suffices to say bioscientists and computer scientists are taught to work in very different ways, and given the project was heavy on programming and light on bio, I didn’t feel the right framework was being applied. On top of that, the project was not really an ‘essential’ part of the lab’s research, so I never had the feeling that whether I succeeded or failed made a difference.

    I suspect your staff member is also aware he/she isn’t meeting your expectations. I don’t know how familiar you are with his/her context outside of lab, but the issues could stem from something other than lack of experience or motivation. I know I never told the aforementioned PI how severe some of my circumstances were – it seemed like telling the truth would have been more likely to have me terminated from involvement on the project, whereas if I said nothing, there was at least a glimmer of hope that I would be able to succeed. That didn’t work. I was never officially a member of the lab to begin with, so I left without any formal statement that I was done. The issues that made it hard for me to make progress still do. Perhaps the worst part of it all is that, when I left without a word, none of the lab members made a peep or sent and email to ask why I was no longer coming to lab. It tacitly confirms my feelings that the project wasn’t important.

    I have two concrete suggestions that may be of use. First, if you haven’t already, have a frank discussion with this lab member about what he/she feels is holding things back. Within the bounds of what’s legal to ask, I would if anything is happening outside of lab that could be interfering with his/her productivity, be it mental health or something else. I would not worry about making the implication that you’re considering terminating employment: if this situation is anything like mine, the lab member already knows that’s a possibility, and may feel like the only option is to wait for the notice to come. If you feel compelled to give one more shot, set a hard date for when the employment will end if changes are not made immediately. Even if things don’t work, you’ll both know when the uncertainty ends.
    Second, I notice you mention a lot about setting different kinds of goals and projects, but these all seem like output-oriented measures. Perhaps instead of looking at the outcome, start looking at the process. Make sure the lab member is keeping a comprehensive timeline and record of what’s being done. This is how computer scientists tackle large projects, rather than by looking at the product. Productivity is measured by comments, concerns, pull requests, commits, updates, etc., rather than by crossing goals off a list. It may not solve the problem, but at least it might give you different insight into what might be off about this lab member.

    May the outcome be in the best interest of you and your lab member.

    • Thank you for your comment, and I am sorry to hear about this experience you had (especially that the lab members did not get in touch with you after you left). The lab member and I do discuss personal circumstances often (the lab member brings them up, and I inquire delicately as well), so I think the lab member feels comfortable talking to me about things outside of lab that are affecting work. And I try to help. Homesickness was an issue at the beginning, and I include this lab member in all my family holidays and in other family outings and dinners when traveling home is not an option for the lab member. I have moved the lab member onto a new project after the lab member professed not enjoying the first one we outlined together (the reason the lab member came to my lab, actually). I am trying to think about process-oriented measures; that is a great point. We do share an online lab notebook so I can see process, but capturing notes daily has been a challenge.

  3. Don’t be hard on yourself.. it sounds like you did everything you could (and more than most would have) to help this person succeed. Every academic I’ve talked to has had some major duds, whether techs/lab managers or students. And it seems like the highest % duds happens earlier in careers, until you’ve figured out what works for you as a lab manager.

    I will say that one thing that did not work well for me when I was the not-meeting-expectations person (see: Toxic Academic Mentors) was increasingly stringent deadlines, guidelines, and monitoring of my activities as time went on. I think it’s hard to go from more to less freedom in a working environment, and it can turn into a vicious cycle.

    • Thanks! I agree that micromanaging isn’t a great solution, and it doesn’t feel good as a manager (to me) to be constantly checking on someone. It’s an interesting question about whether tougher managing situations happen more in the first years as a PI. On the one hand I would expect that is true. On the other hand I’ve had colleagues tell me the best students and postdocs come at this time – maybe because they truly want to work with a new PI. Interesting to think about!

  4. i had a couple of thoughts about this overnight my time, but since that was during the day your time others have already covered most of them, so I will just add a couple of dot points – not really any practical suggestions though.

    * predicting is very hard – predicting whom to hire, whether a project will hold their interest, whether you will get on together as well as anticipated … and for the employee, even predicting if they will like doing research (particularly as you mention they are ‘trialling’ this type of work before deciding on a PhD). (I found I didn’t actually like lab work as much as I predicted I would and eventually concentrated on pedagogical research, which I predicted I would have had no interest in whatsoever).

    * sometimes people become more committed to something they ‘own’ rather than something they see as peripheral to another project – which refers back to points made by Whychosis, although her experience wasn’t ideal. Then different types of deadlines and outcomes and process markers can be identified. But it seems you have already done this.

    * over many years of different types of employment, and watching others who are happy or not in their work, I have developed a theory of ‘mutuality’ – if you don’t see yourself as a good fit, the other person doesn’t see you as a good fit either. For whatever reason. sometimes everyone has to just buckle down and get on with it. If there is a way out, then afterwards everyone is happier. Not at the time, but on reflection later. From what you say, this person may not really be content doing lab-based research, at least on your types of projects (see ‘predictability’ above). If so, she/he has learnt a valuable lesson about themselves and this trial before embarking on a research career has been a successful trial. (If I am misreading this situation, please ignore that sentence).

    Finally, I think you have done everything I would have wanted any of my supervisors to have done … the worst thing is to be ignored and trying to puzzle out all the problems on your own and feeling that you are not contributing because no-one is interested. And you have done much, much more than just take an interest and give advice. Whatever the final outcome in this situation, don’t blame yourself or feel guilty. You are obviously having success with your other proteges, and by taking on those who are not *yet* committed to doing research projects, on more than one occasion, you are giving them valuable insights into their own interests and skills. While not seeming to be optimal in your own lab, or in allocating your own time, you are doing other people a valuable service whether they realise it now or not. But if you do let them go, it might be worth mentioning this point. Someone said something similar to one of my kids after a work experience program “don’t go and do a PhD unless your heart is in it”.

    So I’ve worked around to a piece of advice after all – when looking for lab helpers in the future, if you want someone who will stay long term, maybe try to ascertain if their heart is in it, or if they are just auditing this type of work.

    • just realised none of my comments answer the question in the title …..’can we measure….’. Perhaps it is such an individual measurement that the only answer is ‘it depends’ … upon many factors including the individuals, their personality, the project and most of all the expectations of the supervisor (and how well others in the lab are meeting those). Seems in this case, you are considering all the alternatives and still you are not satisfied. Perhaps (just a suggestion) it is all relative, no absolute ‘measure’ is possible nor desirable.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s