Let’s face this ugly truth: no one teaches you 90% of the nitty gritty stuff that you need to run a lab. Sure, sure, sure, in grad school and postdoc you learn how to generate a hypothesis, how to design experiments to test the hypothesis unambigiously, how to analyze the data and go on to the next set of experiments, and how to fit all of this into the bigger scientific picture. Maybe you learned how to play together nicely in a larger group … or not… during that time.
But when and where during that time did you learn to lead a group, and lead a group effectively? Where did you learn to manage multiple projects going on simultaneously? Where did you learn to manage people with all different personalities and backgrounds? Where did you learn to decide when to keep the peace and when to fire an employee that is toxic in your group? How did you learn to keep track of money in multiple accounts, budget money, and negotiate for resources?
Let’s start with leadership. I realized some time ago that leading comes naturally to me. I don’t think I learned it anywhere- at least I don’t remember working at it. I remember hanging back during grad school and not wanting to put myself out front too much. I remember returning to vet school knowing that I was different, for starters I was about 6 years older than the rest of my classmates, and embracing that for the first time as a strength. I remember making this transition where I became unafraid of sticking my neck out, perhaps too bold at times. But no one taught it to me- it was more a realization that I came to all on my own.
Over the last year my lab group has grown quite a lot. My first couple of trainees are people who seem to me also to be natural leaders, so somehow I quietly developed the naive assumption that everyone that I hired would have those skills. Naive was the key word there. Turns out I have a trainees with a range of different skills in this area. Some of my trainees came from very traditional grad labs where they had a project and they worked on that project in isolation- seems like just the trainee and the PI were involved in these individuals thesis projects. In these cases I think it is safe to say that such trainees haven’t had the opportunity to lead in the way that you have to lead when you are a running a lab, and if leadership isn’t something that comes naturally- then it is something they have to have the opportunity to develop. Some people will probably say that perhaps leadership, or willingness to stick your neck out, isn’t something that SHOULD be taught in graduate school while developing those critical scientific skills. I guess I just think that there is room for both, and I’ll go farther and say that I think one is necessary for the other. After all, the gold standard of all of these NIH training programs is to turn out PRINCIPAL INVESTIGATORS.
Why do I think this is so important? Well, I’ve seen junior faculty come in at various places and have tremendous trouble staffing their labs and managing their staff. You can not make it as a PI without a strong funded research program. You can not build a strong funded research program without staffing your lab, and without developing a training record (or at least it is very difficult). If you can’t staff your lab, where is that preliminary data for grants going to come from???? Leading a lab and managing people effectively is something we right now teach strictly as on the job training the moment junior faculty hit the door of their very first lab- and I think that is a huge mistake.
Blue Lab Coats 
It is a HUGE mistake. Traditionally, through grad school and a post-doc the emphasis is on learning how to ask questions, analyze data, develop papers, grant writing etc. But the idea of managing, delegating, accounting, project management is not rarely touched on. I have seen some universities try to address these issues. I know my current university provides workshops to grad students and post-docs on how to manage projects effectively, how to teach / TA better. Thankfully my PI encourages us to attend them, which isn’t something I can say about other PI’s.
Excellent post and a major issue in the academy right now. A couple of my colleagues are writing an article for the press about this, from a postdoc’s perspective. Kind of a, “For crying out loud, let me learn, and teach me what I *need*”. There is way too much emphasis on extended bench training for postdocs and not nearly enough on *how* to run a lab/project.
tideliar- Better to learn as a postdoc than as on the job training in first faculty position, where the mistakes count and count big. One thing I haven’t discussed is that when jr. faculty make a mistake managing personnel early, and personnel are unhappy about it- word spreads like WILDFIRE. Bad reputations can develop quickly and then you can have trouble recruiting students and otherwise hiring people into your lab. Those first few moves you make with personnel are so critical.
Scientist mother- I agree. Workshops are good if people will attend and if they are taught by people who really know what they are talking about. One of my colleagues mentioned recently that training grants for grad students are supposed to include some component of ‘project management’. I suppose I’m going to have to go read those guidlines now.
I haven’t reached professor status yet, but I have to say the place I most learn leadership as a grad student is with multi-lab collaborations. I’ve worked on a couple projects involving multiple PIs and their students. In those cases the PIs decree what needs to be done, and leave the students to delegate and divide the work amongst themselves. Situations like that are good training, I think.
In undergrad I was project leader all the time, but I work with lots of Type A sorts here, so if someone else feels strongly I much more often will assume a supporting role. I’m trying to learn to be productive and natural in both roles, and I agree that working in isolation doesn’t really hone your skills for either one.
I spend substantial time and effort with my post-docs mentoring them in how to be mentors themselves to more junior trainees in the lab.
A bad trend with junior faculty I see happening all the time is that they get their first big pile of moolah (either startup or grant) and all they care about is hiring minions, to “have people” working for them. It’s a power trip to “have people” in the lab.
A mentor who can teach you how to be a mentor yourself is instrumental. That relationship, supplemented with observing other leaders (some to emulate, others to avoid) is, to me, the way to find your natural leadership style. I also think how well you establish your leadership style before you add in the other pressures of being a PI that you refer to (managing conflict, balancing multiple projects, tracking resources, sustaining the lab’s funding, etc.) will determine whether you can hold onto your mentoring principles when stress is at its peak.
This is a real problem. I grew up in small labs, where everyone ran their own projects. I have never been involved in a large, multi-lab collaboration. I have learned about how to run and lab and be a leader by watching and evaluating the mentors that I have had through my training. I am starting up my lab now and just starting to get it staffed. Right now, I am trying not to make some of the same bad mistakes that I have witnessed over the years. The challenge, for me, is to get folks trained well enough that I can start to give them more control over the daily details. There is certainly a lot of on-the-job training when you move from the postdoc to the PI!
I learned at a National Lab. Taking a postdoc there was the best thing I did for my career. My advisor had me supervise summer students from my very first year there. This was awesome in learning how to supervise people. A staff position at a National Lab is very much like a TT position in terms of learning to balance leading a lab, competing for funding, and doing service. I learned on the job, and was lucky to have a good mentor to help me figure things out. There is more of a “net” at a National Lab, so the learning curve is not quite as steep, in my opinion. I am still learning how to be a prof, after one year on the job!
I got to mentor several undergraduate researchers while I was in grad school. That helped me out so much for my intended career path (PUI). I can’t managing running my own grad students, however.