Responsibilities of a PI

The previous post reminded me of a document I once wrote about what I viewed as a PI's responsibilities. I just reread it, and it's quite a long list of responsibilities. It reminds me of why even thinking about trying to be a good PI is very intimidating. I've copied and pasted the document below.

Note: This list is from my point-of-view as a grad student. I realize it's student--centric. And I realize I've left off many professor responsibilities, such as teaching courses, university service, and grant and paper reviews. Even without those responsibilities listed, the list is looooooong.

What a PI should do

Get grants

Manage money ethically and fairly

Hire good students/postdocs/people

Have good ideas for projects

Manage the lab and people in the lab
-Mediate conflicts of interest/personality between lab members
-Minimize said conflicts by having good and fair policies in place

Ensure safety of lab and lab members

Make sure people get adequate training
-In the essential skills needed to do research
-In the essential skills in their field
-In the essential skills for their desired career

Give people a reasonable amount of intellectual freedom

Not abuse power
-Not ask students/postdocs to do PI's personal errands

Fairly divide non-research tasks (e.g. equipment maintenance, group meeting scheduling)
-e.g. make a list of non-research duties with approximate time commitments
-assign lab members duties so time commitments are reasonably fair
-if can't be fair, draw from a hat and trade every year or six months
-maybe senior students can have a break in responsibility, but otherwise, no favoritism

Not overwhelm lab members with trainees
-Have a plan for each additional lab trainee
-what project will they work on?
-who will train them with what they need to learn?
-ideally, they will receive training by joining a project they can help with
then their training is targeted, and their trainer gets something in return (help)

Give students and postdocs projects that will further their career
-*publication quality projects*
-not repeats of stuff some other lab did that may or may not be hard and time consuming
-not anything that isn't linearly related to their individual publishable project
-not favors for other people/projects they will not receive credit for
-if necessary, should be an explicit trade of favors between individuals
-not technician work (other than above mentioned non-research tasks, fairly divided)

Ensure people get credit when credit is due
-Manage projects so authorship is as clear as possible
-who will be first author?
-what is required to earn authorship?
-ensure people know if their contribution is a favor, not a contribution yielding
authorship

Introduce lab members to people in the field
-Talk up the lab members.
-Their career will impact your career. (i.e. their career good = kudos for you)

Send students and postdocs to present at conferences
-They need the exposure and network opportunities
-You also need exposure of their projects
-Again, their career good = kudos for you
-Help them network at these conferences

Give trainees expectations
-Vacation time expectations (grad school is long, they will need vacation)
-Hours and/or productivity expectations
-if you're happy with their productivity, you won't care about hours
-if their project is not producing, of course you'll want to see their commitment via hours
-years/publications required for leaving/graduating

Give trainees feedback
-What are their strengths?
-What are things they need to work on?
-What are their career goals? What do they need to work on for those goals?

Get trainees feedback
-What can the PI do to help trainees be productive?

Stay high level and big picture as much as possible
-You want good science getting done, they almost definitely want that, too
-Give them suggestions.
-Kindly argue with them about science (don't insult them, don't deride them, don't threaten them)
-Let them try their own way and make mistakes.
-Only insist on your way as a last resort.
-Amount of independence will change as lab members mature scientifically
-And/or senior lab members may direct junior lab members at a more specific level

Advise trainees on what is needed for publications, thesis and conference presentations
-The work will always be "in progress" (the beauty of science...)
-The work can always be better
-You can always figure out more stuff
-When is it enough to write up or present?