The exhalted Goddess Herself Isis posted a poll for the masses asking how often PIs should meet individually with their trainees, divided into grad student and postdoc levels. As implied by her initial separation, the poll respondents feel PIs ought to meet with their grad students more frequently than their postdocs. Of course that's not surprising. Presumably you've learned something as a grad student. Though, I've seen enough in my time to know that's not 100% correct. It's also notable that people seem comfortable relying on the postdoc to bring issues to the PI's attention.
All of my PIs up to this point in my life (5 of them to be exact) have generally been of the "when they have new issues/data to dicuss" variety. Only my very first lab job, 17 years ago (*headdesk* that f'ing long, really? Seriously?) was more directly supervised. Perhaps a bit too supervised, now that I think of it. Let's just say you shouldn't walk up behind me, rub your hands together and say, "So are you at loose ends?" You're liable to get a glass pipette stuck in your neck.
Every other person has had a hands off approach. In fact, I have more formalized meetings with my current PI, though I suspect that's simply a function of sheer lab size. But that's works for me. Gimme some time to try and figure it out on my own, to muddle through for a bit. Chances are I'll get there. Sure, I might waste some time in the short term, but it will be something I never forget. As the old saying goes, nothing instructs like failure.
Classic example: Nat is a fresh faced graduate studet (well part deux grad student, more on that later perhaps). I was recording currents from nociceptors, and I was giving multiple stimuli with the intent of averaging the currents. You know, reduce the noise by 1 over the square root of # traces, that sorta stuff. But the currents were changing rapidly sweep to sweep, which made the average currents look weird to my naive eyes. I thought it might be the solutions, unhealthy cells. whatever. I spent a few weeks trying some different approaches, but never really got anywhere. Finally I went to my advisor. His immediate response, "Well, increase the interstimulus interval, and it's ok to just average fewer sweeps." And of course that worked like a charm. The currents weren't quite as pretty, but they were more meaningful. And it let me get on with the intent of the experiments.
You might say I wasted those few weeks. But there were important benefits. First, it seared the importance of stimulus rate into my brain. It's something I'm always cognizant of in my own experiments, or in critiquing others'. And secondly, it led to two projects that explicitly examined why there was so much use dependence to the currents in the first place. Turns out the sodium current and the potassium currents both have a lot of inactivation, which strongly influces the cell's firing pattern. So really that time was only "wasted" in a very narrow analysis.
On the other hand, you don't want to let your trainees flounder. The fine art of balancing benign neglect with overbearing control freak is why you PIs get paid the big bucks, amirite?
Speaking of control freaks, who are these assholes saying that post docs should meet weekly with their PIs? What the hell for? "Well, you've been here another week, how much closer are you to that Nature paper? You're gonna need that to get job you know." Yeah, that's pretty damn helpful. I guess that's what I've been missing all these years. There's nothing for the creative spirit than being reminded of your failures weekly. Hell, let's make it thrice weekly. You wouldn't want any shred of self confidence to stick around, right?