Monday, June 01, 2009

Postdoc Mentoring

When I was a postdoc, I was just happy to get through a day without being groped (by an emeritus professor), excluded from using the research facilities I needed (by technical staff), yelled at (by office staff), unnerved (by a large male grad student who frequently expressed the opinion that 'girls like to be hit'), insulted (by one of a wide range of people), or the target of a scary lab prank (by one particular technician). The concept of 'postdoc mentoring' was not even a gleam in anyone's eye. I did my work and got out of there as soon as I had the opportunity.

That said, during my postdoc I made some lifelong friends, I discovered that I was good at research, and I learned how to keep going in the face of adversity. I dealt not only with the above-listed items but also the possible tanking of a research project owing to a colleague's reneging on part of the research. I was left to fend for myself in terms of research projects, so I used some of my own ideas to create a project, carry it out (in part by traveling to other universities to do the work), and publish it. It was an important experience for me, and I emerged from it angry but confident.

Would my experience have been more positive if my supervisor, an entirely decent if somewhat clueless person, had been required to at least contemplate 'mentoring' me and assisting me with career development skills?

Starting this year, NSF proposals that request funding for postdoctoral researchers must include a statement about how the postdoc(s) will be mentored. For a brief time this mentoring statement was supposed to be part of the body of the proposal, but, perhaps in response to complaints, the postdoc mentoring text is now a supplementary document, up to a page in length. As with the required Broader Impacts component of proposals, NSF is serious about the mentoring statement: proposals that request funding for postdocs but that do not contain the mentoring supplement will not even be reviewed.

The proposal guide lists the following as examples of mentoring activities:

1. training in preparation of grant proposals, publications and presentations;
2. career counseling;
3. guidance on ways to improve teaching and mentoring skills;
4. guidance on how to effectively collaborate with researchers from diverse backgrounds and disciplinary areas;
5. training in responsible professional practices.

I am trying and failing to imagine my own postdoc supervisor giving me guidance on mentoring skills or responsible professional practices, so I think I will move on and consider instead my experience and philosophy as a supervisor of postdocs. How am I doing with respect to the listed items? In fact, I'm not doing so well, although I think overall I am a decent supervisor of postdocs.

Item #1. For me, this is the easiest one to accomplish. It is difficult to imagine a reasonably sane and functioning postdoc in my research group not getting a lot of experience with these activities in the course of a typical postdoc. I certainly work closely with my postdocs on writing papers and preparing presentations, and there's typically a grant proposal in the works that can involve a postdoc interested in such things.

Item #2. I'm not exactly sure what this means but I know that the only career counseling that I do is of the informal, conversational sort. My own experience begins and ends with academia. I can speak at length about academic jobs and how to approach acquiring one at various types and sizes of institutions based on my own experiences in several different countries, but career counseling about industry, government, or other career modes would have to come from someone else. There are, however, workshops and conferences and colleagues with this expertise, but other than pointing these out to my postdocs (who are generally more aware of them than I am), I don't have anything compelling to say about this possible item in a mentoring statement.

Item #3. My postdocs, if they so choose, can supervise or help supervise undergraduate and MS students, but they typically do not teach. Perhaps I am failing to provide my postdocs with the necessary career skills they need to succeed in a faculty position, but those who want teaching experience during their postdoc either participate in some workshops that prepare grad students and postdocs for the various components of a faculty position, or they acquire a visiting assistant professor or lecturer position before or after the postdoc.

In a short (1-2 year) postdoc in particular, there isn't much time to do anything except the research. Most postdocs enjoy having this time to really focus on research. It may be the one time in an academic career (other than the occasional sabbatical) when you are free of taking/teaching courses, taking/giving exams, and doing endless managerial and administrative tasks. On the one hand, research-only experience may not prepare you for a faculty position in which you have to balance teaching - research - service, but it can set you up well for the research component if you start some long-term projects and develop important collaborations that will carry you through the first few crazy years of a faculty position.

Item #4. By this point in the list it is clear that my mentoring skills -- at least according to the items listed by NSF -- are not as organized or complete as they could be. In my opinion, the best way to effectively do item #4 is for the postdoc to work on a research project that involves a diverse group of other scientists. Most of my projects are diverse in terms of disciplines involved, most involve international collaborators, and I suppose I add a splash of diversity as an FSP.

I'm not sure what to do about the word 'guidance' for #4, though. I probably rely too much on the 'lead by example' type of passive 'guidance'. If I saw a problem with how a postdoc interacted with another scientist owing to their being from a different field, country, ethnicity, or gender, I would certainly leap into action, but other than that, most things get figured out just by working together and doing the research. I don't think this philosophy would sound very impressive (or competent) in a postdoc mentoring statement, and it may well be an example of the flawed philosophy that resulted in the need for such statements.

Item #5. Well, there are certainly a lot of opportunities for this at my university. I am required to participate in them from time to time as part of being allowed to be a PI on grants, but I have found every single one of them without exception to be a huge waste of time and largely irrelevant to my experience as a professor of the physical sciences. I would not voluntarily subject my postdocs to these ethics training sessions.

Here again I prefer to lead by example and discuss informally issues related to co-authorship and credit and sharing/stealing ideas and so on, but once again I don't think that would look good in a mentoring statement. I suppose I could still list the possible opportunities for ethics training. Would that be ethical if I had no intention of requiring or even encouraging a postdoc to participate in them?

What else could be on the list for mentoring activities? What do postdocs want? (other than higher salaries, better benefits, and, in some fields, more respect). Perhaps I am lacking in imagination about this because I am currently interacting with extremely happy postdocs. In fact, my postdoctoral supervisory experience has either involved happy, productive and energetic postdocs or deeply dysfunctional insane and/or unproductive postdocs. If mediocre postdocs exist, they have not come to work with me.

I am glad that NSF is taking postdoctoral experiences seriously, not just in terms of the research but in terms of the overall experience of postdocs and at least asking PIs to contemplate career development issues with respect to supervising postdocs. Perhaps just recognizing the importance of these activities is a major first step towards realizing the goal of improving postdoctoral experiences.

That said, I hope that NSF will cut me some slack if I write a rather lame statement and will consider my track record of postdoc supervision.

If any readers have already written one of these postdoc mentoring statements and is willing to share it (or a draft), please send it to me by email and I will post some/all of them. In addition, postdoctoral readers should feel free to add to the official list of NSF items above; what else should it include?


Anonymous said...

if what is meant by 'postdoc mentoring' is helping postdocs 'succeed' then I think that what you covered is necessary but not sufficient.

In this day and age, being excellent at research, teaching, writing and grantsmanship (and you don't learn granstmanship from being involved in just one grant as a postdoc) is still not enough because the competition field is so huge. the postdocs I knew who were successful in getting tenure track jobs got them because their advisors helped them get their foot in the door - introducing them to their powerful colleagues, putting in good words for them when it was heard through the grapevine that such-and-such department at such-and-such university is thinking of hiring new faculty, influencing hiring committees through other political string-pulling, adding the postdoc as a co-investigator on your own grant so they look like a better faculty candidate if they already bring funding to the table under their own THAT is what 'postdoc mentoring' in the 21st century is about!

And if it seems that this is not really mentoring but is doing favors for the postdoc...the reality is that many new faculty these days are hired based on exactly that, so by not doing it too you are actually putting your own postdocs at a disadvantage compared to their competition.

it is also true that the majority of advisors do not do those 'favors' for their postdocs. But again, from what i've seen from being on hiring committees, the postdocs these days who do get faculty jobs tend to be the lucky ones who have one or more of those favors being done for them (in addition to being highly qualified, which is already a given due to the large size of the applicant pool for a single job opening)

James said...

In your experience, what are the differences b/w mentoring a male postdoc vs. a female postdoc?

Or just depends on the individual, what are your general observations b/w males and females (new PhDs)?

Anonymous said...

I think something that can make a real difference to post-docs (and would be a nice concrete example for NSF) is encouraging your post-doc to go to relevant conferences / workshops / summer schools rather than always keeping them in the lab. Meeting new people is going to give a post-doc a chance to 'collaborate with diverse backgrounds' and 'build their career' etc. I was lucky in both my PhD and post-doc to have supervisors who pushed me to go to relevant workshops / summer schools and at every one I found a useful connection (the next job, a new collaborator, even my husband!). I knew other junior scientists who were never allowed to go to meetings and their careers suffered for it.

Anonymous said...

I think the idea is that good mentoring is more active than what a lot of supervisors do. I've known postdocs who need very little mentoring, because they already have it all together, and others who stumble through and do okay but would have had a much easier time with some active guidance. Not on the level of a student, but someone checking in and bringing up these subjects. I think for some of those bullet points, saying you would have informal conversations is fine, because conversation is a big part of mentoring. The trick is to actually bring up these subjects yourself, instead of waiting to see if the postdoc has questions (and the nerve to raise them). Ask if they have career plans and an idea of what they want to do next. Ask if they want to help out advising an undergrad for you, to get some experience. etc. etc. Tell them about the crappy training sessions (it's a shame they aren't very good at your university). It's not particularly more than you are probably already doing, and it's not even bringing these things up all the time. It's more putting the subjects on the table as something the postdoc can come to you about if they have questions, and bringing them up to make sure your postdoc is thinking about them.

Candid Engineer said...

This sounds very similar to the "training plan" portion of the NIH postdocotoral fellowship application package. The PI is supposed to talk about all of the ways that s/he and her/his lab will provide an excellent training environment for the postdoc, how the experience will forever enrich the postdoc's life, etc, etc.

My "training plan" section consisted of various paragraphs discussing the following:
1. expected/encouraged collaboration within my current group, with outside medical professionals, and with industry
2. learning new experimental skills that are very different than what I knew previously; these skills will "synergize" to create a marvelous independent researcher, blah, blah, blah.
3. participation in local/national/international conferences; PI exposes postdoc to his/her professional network; postdoc has opportunity to make great connections
4. I included a paragraph about the specific things my PI does to go out of his way to mentor postdocs (e.g. we have a bimonthly seminar series focused on topics such as 'securing funding', or 'starting up a lab').
5. PI will do everything s/he can to ensure excellent job placement for the trainee, from providing advice about applications/interviews through negotiating an offer.

Hope that helps.

Professor Staff said...

This sounds similar to the "training plan" required for NIH NRSA postdoctoral fellowships. I've reviewed these before (haven't written one - I've mostly hired foreign postdocs who are not eligible), but others here may have some to share.

It may vary by study section, but the one I have served on took them quite seriously, and they were sometimes 3+ pages long. Personally, I felt that most of the time they were critiqued by what they did NOT contain, as opposed to the highly unusual rare case of something novel in this component.

Anonymous said...

This has been part of NIH postdoc applications for a while. Unfortunately, I feel like it's one of those sections that does more to benefit established labs than anything else. You just need to know the bullet points they want to hear and fill them in. If your PI knows how to play the game you have a chance at the grant and, if not, it's makes sure the grants go to other labs.

I wouldn't be so cynical, but this is one place where there's no way to check on follow through. The research section can be compared to later publications, but there's no way to ask if someone did what was said in the mentorship section.

That said, it's nice to require it in grant applications to force PIs to think about mentorship (especially now that only 2 year postdocs are less common), but I wish it wasn't weighted as highly.

Ria said...

My postdoc mentor spent quite a bit of time providing his postdocs with the means (a technician and an undergraduate student or three) to learn mentoring and personnel management, while also learning to delegate (when you're managing at least three full projects, you can't do the work all yourself!) in preparation for the rough adjustment to non-benchwork PI-dom. We were strongly encouraged (I would almost say "required") to submit at least one grant for postdoc funding if we wanted to work in his lab, and were also encouraged to continue to write grants, as time permitted. All in all, I worked harder than I have ever worked in my life, created enough data to analyze for years, got three first author publications from 3.5 years of postdoc (not to mention at least 6 other publications), and learned a whole host of skills that are obviously important to succeed as a PI, but aren't valued in the initial determination of whether or not you get to BE a PI.

All in all, I would say that he managed to show more forthought in his postdoc mentoring than is typically indicated by the stories I hear from others.

Anonymous said...

As part of the Bio-side (translation - NIH), I'm glad to see this happening at NSF. NIH tends to lag behind but at least this may make it a part of future R01 type grant applications. There are far too many 'kingdoms' where post-docs are being used inappropriately to gather more and more data and not being guided to a future career. At a recent women-in-science panel, adding this component to NIH grants (for grad students AND post-docs) was one of our suggestions that I think actually is possible (we can all wish for better family-balance, daycare, etc)

One thing you can mention - how many conferences will you support this post-doc to go to (provided they have data).. are these conferences good for networking? For those with non-academic goals, are there more appropriate meetings? Also, as a senior person, you can mention the track record of your previous post-docs. I think it's fine to mention the various seminars about ethics, etc even if you aren't requiring them. As a post-doc I took advantage of various panels/seminar series just to learn a little more about things. (again, as NIH-side, we ARE required to do the ethics courses, and yes, they are pretty worthless).

Astronomum said...

In fact, my postdoctoral supervisory experience has either involved happy, productive and energetic postdocs or deeply dysfunctional insane and/or unproductive postdocs. If mediocre postdocs exist, they have not come to work with me.I find it very hard to believe you have only had happy postdocs (apart from the crazy ones) - maybe they were hiding they're unhappiness from you. I have yet to meet someone who is truly happy with the postdoc experience. The constant worry over your future in the field and the expectation that you work all hours of the day just seems to drive people down into the ground. Most of the time we postdocs ignore this and just get down to getting some work done - but it still doesn't make us happy with the situation... I would never say this directly to any supervisor though...

I think it's great that NSF realises something needs to be done, even if they're list does seem likely to not quite reach the heart of the matter. Frankly I suspect that no amount of mentoring will improve a career model that was set up only to deal with young men with (usually non-working) following wives, and has not yet adjusted to realities of modern life.

In most other professions it seems like by the time you're in your early-mid 30s you are a respected "member of the club" (even Doctors who surely need as much if not more training than research scientists). Why does research science support a career path which has the average age of a first (non-fixed term) job starting to approach 40...?

Candid Engineer said...

@ Astronomum:
I find it very hard to believe you have only had happy postdocs (apart from the crazy ones) - maybe they were hiding they're unhappiness from you. I have yet to meet someone who is truly happy with the postdoc experience.Maybe it's not common for postdocs to be happy, but it's far from unheard of. From what I can tell, FSP runs a great lab and so why can't her postdocs be happy?

One and a half years into my postdoc, I am as happy and content as I've ever been in my life. I am in a terrific lab, I work 45 hour weeks, and I get my shit done. Many other postdocs in my lab operate the same way and are equally content.

The postdoc culture is one that is very field-specific. I'm not sure what astronomy or FSP's field is like, but in my field, you are *not* expected to bust your proverbial balls to land scads of fascinating data and publish fifty trillion papers. In my field, it is more common to do postdocs to learn new techniques and get different research experiences, and with the help of a supportive PI/research lab, it is more than possible to lead a productive and happy postdoc career.

Small college science prof said...

I don't doubt that NSF takes this new requirement seriously, but do you have any indications of how seriously the reviewers will take it? I ask because there's often a disconnect between NSF's expectations and reviewers' actions. For instance, I have had proposals panned by reviewers for being too expensive, even though my NSF program officer told me explicitly that reviewers are supposed to judge only the science, and it's the program officer's responsibility to judge the budget. Also, I know NSF takes the Broader Impacts statement seriously, but I myself as a reviewer have never given much weight to that -- I'm just looking to see if the proposed science is good. Have you had reviewers comment specifically on the postdoc mentoring section?

Kevin said...

I've not had much experience mentoring postdocs.

After being a professor for 26 years, I've only ever had one postdoc (postdoc positions are still rare in computational fields---a lot of students go directly from PhD to industry or faculty positions).

I did not have funding for him as a postdoc (I paid him out of a grant by not taking summer salary and not buying computers we needed), so most of his time was taken writing grant proposals---none of which got funded. He finally took a funded postdoc in England with a friend of his.

I'm now mainly trying for NIH funding, and NIH loves postdocs and hates grad students, so I'm probably going to have to figure out how to mentor postdocs. I think I've got down how to mentor grad students at least moderately well, so could someone explain to me what the main differences are between postdoc and grad student mentoring?

Janus Professor said...

I spoke with one NSF director who told me that they wouldn't fund any proposals with funds requested for post-docs! This director said that the mission of NSF was to educate, and they took that quite literally to mean graduates and undergraduates. Glad I asked!

Ms.PhD said...

A few things-

First, I can entirely believe that FSP has only had happy postdocs. Most postdocs are happy the first 1-2 years, even in Bio where things are generally crappy. Even I was pretty happy with my first 1-2 years of postdoc.

Second, FSP your list is pretty good. I would just say a couple things.

a) You should stay on top of the listings of what is available nearby for your postdocs. You shouldn't expect them to be aware of this themselves, and knowing what's available will also make it obvious to you when there are gaps. Two of the biggest issues for us as postdocs is that many of these seminars on career things are only offered once a year, so if you miss one the next might be too late. And the other is that sometimes they are discontinued, especially if they are run by the postdocs ourselves, it's hard to keep these things going because there is so much turnover. Finally, it might be really good of you to make a point to speak at a few of these things! We need more people like you to show your face - and you might learn a thing or a few.

b) Informal isn't really good enough, in my opinion. My advisers have done a great job of claiming they would advise me, at least in writing, or asking me to write this section for them. But they do not follow through, even when I ask them to.

I would be really impressed to see a PI who said "we will discuss at our weekly meeting x, y and z" and monthly I will meet with the postdoc specifically to discuss "abc" and annually to discuss "d".

That sort of thing would be amazing, especially if PIs really made a point to stick to it.

The savvy postdoc will ask, too, but if you're both really in it for the win, then you'll do all of that plus what the first Anonymous said, and pull some "favors" in when needed.

Sally said...

The astronomy decadal survey is in progress, and this white paper on Training the Next Generation of Astronomers is getting a lot of attention. A major theme in the paper is how demoralizing the postdoc phase has become.

Schlupp said...


Interesting, given that about 80% of the postdocs I know in my field and in the US are paid from NSF grants. Also, it seems strange that they would invent a postdoc-mentoring supplement if they decide to discontinue postdoc positions.

Anonymous said...

I think it is important to remember that many postdocs simply WILL NOT KNOW many things that seem self evident 2 or 3 or 4 years down the track. For example, someone who has done great PhD research may have been so focused on the research, they may not realise how competitive it is to get a job, and how important it is to publish in a timely manner. This may seem improbable, but I can assure you I know people like this. They may also not know a good strategy to network at conferences to start with, or how to write a cover letter, or how to promote themselves.

Good mentoring also means simple things such as an advisor reading manuscripts in a timely manner. You might be extremely surprised how often a mentor sits on a manuscript for long periods - eg over a year (it has happened to me)- and a postdoc allows that to happen because they do not yet have the skills to progress the situation, or realise how critical it is to get that paper out.

And I think offering a postdoc a small amount of teaching, even 3 or 4 lectures, can be useful for them.

Anonymous said...

A common theme in the grad and postdoc labs of myself and my colleagues was the pattern of no communication from the advisor for long stretches of time (because the advisor was busy with other things) or not giving much thought when asked for input (again due to lack of time) and thus allowing the student/postdoc carry on a certain path for a long time. or actually telling the trainee to carry on a certain path.

Yet much later on, after the trainee has invested much time and effort on the path that was agreed upon, now when the advisor is FINALLY ready to pay attention to that particular student/postdoc, he/she is then displeased at what the trainee has been doing all along, not because the experiments were done incorrectly or the data has taken the project in a new direction, but simply because the advisor didn't realize until now what he/she really wanted because of being too busy to think about it earlier. So the advisor changes the topic or research direction, doesn't stay abreast of new developments and so the cycle repeats.

Constantly moving the goalposts on the trainees is very frustrating for them and slows down their progress, and it's even worse if the trainees also get blamed as if it's their fault they didn't know this is what the advisor was going to decide now that he's actually had time to think about what he wants.

In industry, managers would be fired for such inefficient use of resources and time. Professors need to stay on top of all their students and postdocs' work IF they are not going to be giving complete free reign to those trainees to do as they please (which would be a different model of mentoring, in which case constant moving of the goalposts shouldn't be an issue in the first place).

mixlamalice said...

As a French post-doc in a US university, I would like to give a few comments (and to apologize for my written English):
- Like Astronomum, I am not particularly happy as a post-doc. This has nothing to do with any French/American cultures clash or any problems with my advisor. This is just that I am almost 30, I still don't know when I will be able to have a "real" job, e.g. something longer than 1-2 years that will allow me and my girlfriend to live together in the same place for quite a long time, and possibly start a family. I'm kind of bored of living like a student.
- I also think that the short term of the post-doc is kind of annoying in terms or research: if you want to work in academia, for your own good you need to add value to your resume, so you have to publish, so you have to work fast and not take too much risks (if nothing works for 2 years, you're basically screwed). I don't think that's the best position to do really interesting and innovative research.
- This time, as a French, I think that contrary to what FSP said, there is a lot of boring administrative stuff that was given to me over the last 18 months compared to what I would have had to do in a French lab. Maybe it is specific to my department or my advisor here, but I have the impression that I spent half my time writing parts of proposals, looking on the internet for new tools wanted to buy for the lab, fill paperwork (perhaps it is because I am a foreigner), and do crappy preliminary experiments about possible collaborations that actually never happen...

Anyway, congratulations and thanks for this interesting blog.

Anonymous said...

In response to Anonymous: A postdoc should not need his mentor to hold his/her hand in research all the time. This is quite unacceptable for graduate students (except, perhaps, first-year students), let alone postdocs.

Dr Spouse said...

I did two postdocs (and a few weeks of a third, we won't go into that). In the first one, a more traditional existing-project postdoc, I had plenty of guidance on collaborative paper writing but there was no suggestion I'd get involved in writing grants. It was cross-disciplinary, though, and I did learn a lot (mainly by example, and some by negative example) about collaboration; some of it has helped me to protect myself now as the only psychologist on medical projects (as that's what I was).

The second postdoc was a fellowship and it was basically assumed I already knew how to write NIH grants and NIH format (I still don't, really, but it was in the US and I don't work there most of the time, now). I was doing my own research, and any collaborative writing was with people outside my institution. I had loads of really interesting theoretical discussions with my mentor or with people she suggested I contact, and I think she saw that as her main job as a mentor.

Career advice is something I take seriously for PhD students and I have always felt my colleagues and mentors have taken it seriously with me too.

Teaching has more been a case of "you can do some teaching if you like; so-and-so needs someone this term". At my current institution this is also taken a bit more seriously (cynic alert: because we need people do to the teaching).