Sunday, June 13, 2010

Can junior PIs make decent mentors?

Following from a discussion on last weeks' post about the new NSF is borked forum, the comments moved towards the topic of junior PIs and whether they should be postdoc mentors. The start of the discussion was sparked by Dr. Girlfriend, who made the comment:

I honestly do not believe the average new PI has the experience to qualify as a suitable postdoc mentor.

I took issue with this being an unsuitably broad statement to make and then we were off on a tangent of no return. So, I thought it might make for an interesting broader discussion. Do you, dear readers, believe that a pre-tenure faculty member can make a good postdoc mentor?

As full disclosure, I obviously have a horse in this race and may or may not be currently in this role. But certainly my comments can be interpreted from the position of someone who feels they can be an effective mentor at this stage of their career.

Most importantly, however, I think it is key to recognize that effective mentorship does not only mean one-on-one activities. As I stated in the previous thread:

We are also making the assumption here that the lab PI is the only person to whom a postdoc can go for guidance, and IME, that is also far from true. As a postdoc I consulted several PIs, both at my home institution and elsewhere, on a variety of different issues from applications, to funding and taking a position. I'm not sure how that would change based on the experience of the primary PI.

It is ridiculous to impose a requirement of tenure on anyone who wants to mentor a postdoc (as Dr. Girlfriend seems to want to do), because every mentor is going to have strengths and weaknesses. A postdoc and supervisor need to be able to recognize these 'holes' and find other mentors to compensate for weaknesses of the PI. This is no different from the situation where a postdoc wants to go into an 'alternative' career and must find mentors that will be able to guide them through that process.


As I have stated before, mentoring is about facilitating the transition from trainee to whatever career, for your peeps. If I am mentoring someone who wants to go into industry, I'm going to make sure they find someone in that field to talk to. Why is that any different when it comes to being a faculty member? Why can't junior PIs encourage their postdocs to solicit other information on being a faculty member from people at different stages of their careers? All of us do this all the time and it wouldn't make any sense to not suggest that to our trainees.

So, I guess I'm confused or maybe just used to ensuring that I have a broad base of mentorship and that my trainees do as well. Perhaps I just didn't realize that I need to be a Swiss Army knife of mentorship, when I probably see myself, at best, as a spork.

24 comments:

  1. I think that assertion is quite odd. My experience is that it mostly depends on the individual advisor.
    But assuming that personality and working styles are compatible, I think there are definite advantages to working with a more junior PI:
    (1) a newer PI is often more invested in your experimental and career success - it's much more important for their career than it is for a more established PI.
    (2) If you've joined a lab to learn from your advisor's experience, you actually get to learn from your advisor.
    (3) you might have less freedom early on (it comes with money) but you definitely have more input into what research happens, and what questions are interesting;
    (4) you get more experience in administration and lab management.
    (5) smaller newer labs tend to be more collaborative, so you work on more projects - and therefore end up more papers and with more experiences (and have less of a competitive-post-doc kind of experience)
    (6) Your advisor is more accessible.

    That's my experience anyway. And my publication record is as good or better than friends in more established labs.
    Generally though, in choosing a post-doc mentor I'd say the most important thing is personality and working style.

    ReplyDelete
  2. The advantages Natct listed are quite accurate. Of particular relevance for postdocs (as opposed to grad students) is point (4), that you'll see a junior person setting up shop, negotiating for an extra room as the lab expands, navigating getting funding in the early years--all things that will be very important to you in the early years of your own faculty job, if that's your plan.

    I'm headed to postdoc with junior faculty and couldn't be more excited about it.

    ReplyDelete
  3. I'm also a pre-tenure PI, but I think that I would be a good mentor for a postdoc. For all the reasons that Natct mentions. But, there is one glaring deficiency of ANY new PI (especially if new postdoc is wanting to stay in academia and get on the tt): track record. This is an important component of most post-doc fellowships, and having one on your CV is important. Not to mention the fact that on a job search someone from a "famous" lab will often get extra attention. I don't think these are insurmountable difficulties, or that folks wanting to land a tt job should automatically avoid new labs. Nevertheless, I make sure that new postdocs know what they are in for so they can be sure to address these "deficiencies" from day 1.

    ReplyDelete
  4. I am a senior postdoc looking to soon transition into that first position, and I think I would have a very unique experience to mentor a postdoc on the realities of this transition, as I have just experienced it. I agree that seeking out more senior PIs for some aspects of your training may be good, but as someone who has been advised by two very senior PIs in charge of large successful labs, they sometimes lose sight of both the realities of today's job (and granting) market as well as the validity of other career paths. Thus, in some ways, a junior PI is an excellent mentor. I have found myself seeking out junior PI's in my department for job seeking advice, as well as scientific discussions. Plus, not everyone (junior or senior) has what it takes to be a great mentor thus those who have that talent need to be utilized.

    ReplyDelete
  5. A new PI has research experience, and if a postdoc is aspiring to be a lifelong bench scientist then they might be ok with that.

    The best mentor for a future faculty is someone who has already succeeded in that role.

    A postdoc should already know how to do research, and needs only experience in this respect.

    What a postdoc needs to learn is how to develop a research program, write grants, manage people and resources, and how to hit the ground running when they get their own lab. A postdoc needs a mentor who has served on search committees and can give advice on what makes a strong job candidate.

    Ideally a postdoc mentor should continue to be your mentor once you get your own lab. If you chose a mentor who has not yet got tenure herself, you run the risk of missing out on a key mentor at that critical stage.

    A postdoc will need to develop a project to take with them when they set up their own lab, and few new PIs can afford to let a promising project leave the lab.

    An assistant professor can be a good mentor to a postdoc, but as a primary mentor they are simply not qualified. A new PI has experience of being a postdoc, but postdocs do not need to be trained to be postdocs – they want mentoring on how to not be career postdocs!

    ReplyDelete
  6. All the reasons mentioned Natct in the first comment are reasons why a new lab is the best option for a graduate student.

    However, a postdoc does not need to be hindered by a hands-on PI looking over his shoulder. Part of being a postdoc is bringing new techniques to the lab, and keeping the lab update on the latest technologies and techniques. A new PI will take that role away from you because they are the best pair of hands in the lab – not you the postdoc.

    A new PI is learning as she goes along, and has no more experience than you in how to run a lab. A postdoc might find that she is better at training students and managing the accounts than her PI – that is not a good place to be.

    ReplyDelete
  7. The best mentor for a future faculty is someone who has already succeeded in that role

    Not necessarily when that success was launched three decades past when the circumstances were considerably different.

    ReplyDelete
  8. Of course junior PIs can make good post-doc mentors. Do they all? No, but neither do all senior PIs. I do have several friends who joined brand new labs and a rough time. If you're one of only two trainees it can be great to have so much attention from your PI, but there can also be lots and lots of pressure because the success of the lab depends so much more on you than it would in a larger, more established lab. And when it's just you and your PI day in and day out, well, there can be tension.

    There is also the issue of age. Especially if you took a few years off before grad school, or are going to do a 2nd post-doc, it's likely you and your PI will be very close in age, which can make the boss-employee relationship a little bizarre.

    All that said, there are several new, young faculty at my institution who have become "side mentors" for me in the last few months. They are awesome, and are full of information and insight that I can't get from my very senior actual PI. One of them is a year younger than me, a fact that induced a medium-sized panic attack when I first heard it, but I'm cool with it now.

    In the end, like with anything, whether or not you are properly mentored as a post-doc is going to be a combination of your own initiative, smart decisions, and a lot of luck.

    ReplyDelete
  9. I think everybody is really in agreement here. I don't see anybody disputing that a postdoc should ideally be exposed to the influence of several people at different career stages, regardless of who is primary mentor, and any primary mentor is effective only if s/he facilitates this exposure. You have to have someone junior who can give you 1-on-1 attention and someone senior who's been around the block. With junior faculty you build a lab from scratch, but they don't have too much weight on your CV; with senior faculty, the weight is there, but often there is little 1-on-1 interaction and mentoring.
    I hired my 1st postdoc near the end of my TT and I think it worked great.

    But let me reiterate another point from last thread: regardless of what we think, the funding agencies are biased (at least in my field) against TT faculty advising postdocs. For my NSF CAREER and one of the DOD young investigator awards, in both cases the PM's told me they don't want me to budget for postdocs because junior faculty are believed not to be experienced enough to mentor a postdoc. So whatever we think, the funding agencies may have their own take on the issue (I hope theirs has some solid statistics to back it up...)

    ReplyDelete
  10. the funding agencies are biased (at least in my field) against TT faculty advising postdocs

    This is quite different from my experience, but it might be a field thing.

    ReplyDelete
  11. An assistant professor can be a good mentor to a postdoc, but as a primary mentor they are simply not qualified. A new PI has experience of being a postdoc, but postdocs do not need to be trained to be postdocs – they want mentoring on how to not be career postdocs!

    This is exactly why a new PI CAN be a good mentor for someone who wants to move on: these people JUST did it, in the current system that a current postdoc will need to navigate. Not in some bygone day of 30-40% paylines and old-boys-network-phone-call job searches.

    Saying that an assistant professor is "simply not qualified" to mentor a postdoc is extremely insulting and pretty damn inaccurate. Like others have detailed above, there are disadvantages, but these can be dealt with by anyone capable of walking around a department and knocking on some doors. The advantages of seeing how to launch a lab space, a group and a modern-day research program in real time FAR outweigh the inconvenience of having to establish additional relationships with complementary mentors.

    ReplyDelete
  12. I still do not get how someone who has no experience in running a lab and managing staff can be a good mentor to someone aspiring the this role.

    A new PI can be an excellent mentor for graduate students because they are hands-on and heavily invested in their success.

    However, a postdoc does not need training regards doing research - they require only experience in this respect.

    Getting a job requires a good publication record, but getting tenure requires much more. A postdoc need to learn how to become a group leader and develop a project to take with them.

    An associate and full professors will have current experience of interviewing and the tenure process because they are serving on search committees.

    ***The advantages of seeing how to launch a lab space, a group and a modern-day research program in real time FAR outweigh the inconvenience of having to establish additional relationships with complementary mentors***

    This is a great experience, but most advantageous to a graduate student. A postdoc has only a few years to publish a significant number of papers, and does not need to be hindered by lack of money or organization. A new PI might not get tenure, and you are left with an example of "what not to do".

    A new PI as as much experience of lab leadership as a postdoc and I am not convinced that "learning together" is a good choice for a postdoc.

    It is much better to seek out additional mentors who are new PIs and watch them succeed or fail from a safe distance.

    ReplyDelete
  13. Getting a job requires a good publication record, but getting tenure requires much more. A postdoc need to learn how to become a group leader and develop a project to take with them.

    Can't you see what we are saying: that a new PI is someone who just did these things in recent years, and probably has a lot of extremely relevant advice on how to do it successfully. There is absolutely nothing inherent in being senior that makes it any more likely someone will be able to impart these things to their postdocs. And in fact, a significant percentage of old-school faculty are doing a terrible job of mentoring their postdocs into the next steps of their careers--just look at how many postdocs are having a hard time moving on, and consider the relative % of junior vs. senior PIs that those people have.

    An associate and full professors will have current experience of interviewing and the tenure process because they are serving on search committees.

    Actually, most search committees have at least one member who is a very new faculty member. This is deliberate, because successful junior faculty input on incoming candidates is valued by most departments. As one example, I just served on a search committee for my department when I was just a 2nd year asst. prof. I learned an incredible amount about the academic job landing process, not least because the other senior profs on the committee were candid and open about their assessment of the candidates--in a way that they often can't (or won't) be to their own postdocs going out on the search. I now have that perspective to give to my postdoc, who, in his previous postdoc lab with a senior, 'famous' PI, did not get any career mentorship.

    ReplyDelete
  14. I was active on the previous post and it seems that this topic brings out surprisingly strong feelings. I think that the answer is (and always will be) it depends.

    Part of the issue is what a specific field requires for faculty. There is one that I can think of (it will coem to you if you ponder for a moment, especially if you remember the inital post was in terms of the Canadian funding system) where the average new faculty member has never won a grant and many have not have ever published a paper (not even a co-authored one).

    Other fields in science require 5 to 10 years of post-doctoral training, at least 10 papers (some of which are first author) and evidence of success with funding. In these cases, a junior faculty member is ideally suited to mentor a post-doctoral fellow -- just like the more senior PhD students can mentor the more junior ones.

    My main concern was putting a post-doc into the funding package of a PI who has never published means that both the PI and the post-doc are learning the ropes for how to publish and how to apply for funding at the same time. It's not a fatal scenario but it is one that I worry about.

    PLS clarified, in the previous thread, that his field falls into the category of post-docs and pubs required so I can't imagine this concern applying to him.

    ReplyDelete
  15. this topic brings out surprisingly strong feelings

    Hahaha. Dude, it's blogging. I could post what I had for breakfast and someone would tell me I'm doin it rong.

    Certainly the relative amount of training a new PI has based on their field is important to what they can offer in terms of mentorship and I don't know if there is a "sweet spot" where the benefit is highest for the trainee. My only contention is that branding all junior PIs as ineffective mentors ignores a sea of variability, and as seems clear from the comments, different trainees are also looking for different things out of their experience.

    ReplyDelete
  16. A postdoc has only a few years to publish a significant number of papers, and does not need to be hindered by lack of money or organization.

    Neither does a postdoc need to be hindered by someone whose interests are most aligned with *one* of her many postdocs having a big hit each year and not caring about making sure *each* postdoc has a medium hit.

    ReplyDelete
  17. Uh oh, do I need to fire my postdoc who just started? Should I trade him in for a couple of grad students? When will I be ready oh sage ones, I'm only a first year assistant prof.

    ReplyDelete
  18. I am a little taken back by the emotional response to my views.

    Did any of you PIs come straight from a new lab?

    Do new PIs really think a postdoc is a better option than a good technician? Can you risk giving a postdoc independence at the bench? Can you honestly afford to let an exciting project leave the lab with your postdoc?

    Your postdoc mentor is you most significant one. Ideally she will continue to be a mentor as you go up for tenure yourself.

    At the end, when you are searching new mentor, all you can speak to his past mentees from his previous lab. All PIs have a proven track record regards doing research, but not all have good leadership skills.

    Tenure at least says a person can get a research program of the ground, perform teaching and administrative duties, and write successful grants. A new PI is a big risk.

    I believe what many department lack is good mentors for assistant professors. I would not feel right mentoring my first students if I did not have my PI there to mentor me in my mentoring.

    ReplyDelete
  19. Nice discussion. I started a comment, but it became a post.

    ReplyDelete
  20. I answered in more detail at prodigal academic's post, but I also wanted to note here that being "adamant" isn't the same as being "emotional." I am quite adamant that I think it is incorrect to assume that new PIs won't be good postdoc mentors, however I am not emotional about it. It doesn't make me cry, it doesn't even particularly piss me off, it just inspires me to try to correct the incorrect assumption.

    If you are not yet a PI yourself, you might not be able to see how you (or someone else) could make a good postdoc mentor fresh out of the box. I myself was extremely anxious and worried about this when I started. But over the past two years of being a TT PI, I have gained a huge amount of confidence through experience and have no doubts about my own ability to transmit what I now know about getting onto a positive trajectory towards tenure (which, incidentally, needs to happen within the first 2-3 years, mad scrambles at the last minute are not a good way to do it).

    ReplyDelete
  21. Maybe I'm missing something... but a new PI has probably 4-5 years postdoc experience, plus the whole job search and setting up lab experience - so how is this not enough to be a mentor to someone fresh out of graduate school? Or are we talking about mentoring a serial postdoc on their 7th year?

    ReplyDelete
  22. OMG - I take a few days break from the blogosphere and come back to find ridiculous statements that junior PIs are not capable of mentoring postdocs. Absolute horseshit. And, for the record, I completely agree with Arlenna's statements, particularly this one:

    "Saying that an assistant professor is "simply not qualified" to mentor a postdoc is extremely insulting and pretty damn inaccurate. Like others have detailed above, there are disadvantages, but these can be dealt with by anyone capable of walking around a department and knocking on some doors. The advantages of seeing how to launch a lab space, a group and a modern-day research program in real time FAR outweigh the inconvenience of having to establish additional relationships with complementary mentors. "

    And to address this comment from Dr Girlfriend:

    "Do new PIs really think a postdoc is a better option than a good technician? Can you risk giving a postdoc independence at the bench? Can you honestly afford to let an exciting project leave the lab with your postdoc?"

    Ummmm ... getting a postdoc in a new lab is a bit of a gamble but can be extremely beneficial. DM and PP have both posted on the pros and cons associated with this decision.

    And postdocs don't automatically take their studies with them when they leave to go to faculty positions ... this is decided through negotiations with the PI. My postdoc is working on a study that I designed but that he has tweaked. We have already discussed what he can take with him when he leaves and it is NOT what he's working on right now.

    Seriously, Dr Girlfriend, you are entitled to your opinions but your personal experiences seem to be clouding your ability to recognize that not everyone shares your opinion. Several of us did postdocs in new/new-ish labs and found it to be a wonderful learning environment that has helped immeasurably in establishing our own labs.

    ReplyDelete
  23. I came to the US as a postdoc from a developing country to a lab where the PI was a yr away from tenure evaluation. I was never mentored in such a way as to tell me what the system in the US is like and how does one succeed in academia here. However, I was asked to finish the two projects that were undone in my advisors NSF CAREER grant and NSF was hounding the PI. I did that and got 3 papers and then moved to another postdoc where the PI was just 2 yrs into tenure track and was having a hard tiem growing plants in the greenhouse for a popgen experiment. While there I standardised growing conditions and along the side did experiments that were my main focus of interest and we got 3 papers - yes the PI took authorship on all papers even though they were my ideas and I executed them. Never once was I encouraged or even told about NSF granting system or writing a proposal ... Now about 7 yrs after having left my second postdoc and still being an adjunct I see that my second postdoc advisor has encouraged a grad student to write up an experiment I began in that lab as a DDIG grant and it was funded. So, all you people, let me know what you think of this mentoring.
    My colleague went to the lab of a full Professor and within a yr he was asked to develop a grant proposal which was funded even when my colleague was doing his postdoc. So in comparison to my experience, let me know what you esteemed ones have to say about junior faculty being postdoc mentors. I have seen many a junior faculty in the department where I am at just 'using' postdocs who are trained in lab work to set up labs and such. Infact when I was in my second postdoc I over heard my 'mentor' and another junior faculty who had come for a departmental seminar discussing why it is better to hire a postdoc that a grad. student. SO go figure people. ALl of us will say what we do is the best but no one really cares about the postdocs' upcoming. ANd now NSF REQUIRES a postdoc mentoring plan in proposals. Bull shit! Another one of those seemingly good ideas like the broader imapcts section - once again NSF's way of making sure that those select few will be able to get grants and hire postdocs. I have seen people getting funded over 600K with postdocs while my grant with apparently a really good idea (the PO told me about the panel discussion) and just a third of 600K$ will not get funded because I am a nobody here as I did my phd elsewhere and no one really identifies with me. The PO officers have told me many a times now that I should network with senior scientists if I want to be successful. And I constantly get comments saying that I should consult an expert in the field and get a letter of support. WHY???
    NSF is all botched - it is a club of elite researchers who keep funding each other, each other's students and ...

    ReplyDelete
  24. Anonymous above is pretty spot on. Junior PIs receive more from exploiting postdocs than helping them. I'm a 4th year PhD student and I've seen a half-dozen postdocs come in, set up projects, publish and then get released because the PI's grants didn't hit. The junior PI has very little experience in competing for R01 and R21 funding (NIH) and so the postdoc is basically just a stepladder for the PI to take a grab at funding, kicking out the ladder if things don't go well. With the scarcity of awards these days, a postdoc in a Junior PI lab is playing Russian Roulette with an automatic pistol, hoping it will jam. A senior PI with established funding knows more how to orient research in a way that is palatable to the NIH, where a new PI will usually try to shoehorn in what he or she WANTS to do....even if it's not hot in the field at the moment. The Postdoc, in that situation, is basically giving research for free, since after the postdoc leaves, the PI will keep those ideas and preliminary data for later.

    A new PI should hire a lab manager (MS level) and 2 experienced techs per project. Getting an MS student or two is fine, too. The PI should lead the idea and direction phase and generate that preliminary data, first. AFTER a successful submission, hire 2 or 3 postdocs to each lead an AIM on the project. The core workers (manager and tech) are the persistent factor and keep the lab running smoothly, allowing the postdocs to spread their work on science, not managing the lab and training PhD students.

    ReplyDelete