Saturday, December 13, 2014

The Shockley model of academic performance

I just came across a very interesting post from Brian McGill about William Shockley’s model for why graduate student performance varies so much. Basically, the point is that being successful (in this case, publishing papers) requires clearing several distinct hurdles, and thus requires the following skills:
  1. ability to think of a good problem
  2. ability to work on it
  3. ability to recognize a worthwhile result
  4. ability to make a decision as to when to stop and write up the results
  5. ability to write adequately
  6. ability to profit constructively from criticism
  7. determination to submit the paper to a journal
  8. persistence in making changes (if necessary as a result of journal action).
Now, as Brian points out, if you were 50% better at all of these (not way beyond the norm, but just a little bit better), then your probability of succeeding in your assigned task (which is the product of the individual probabilities) is roughly 25 times better. This is huge! And it’s also to me a reason for great hope. The reason is that if, alternatively, being 25 times better required being 25 times better at any one particular thing, then it seems to me that it would require at least some degree of unusually strong innate ability in that one area. Like, if it was all about writing fast, then someone who was a supernaturally fast writer would just dominate and there’s nothing you could really do to improve yourself to that extent. But 50%? I feel like I could get 50% better at a lot of things! And so can you. Here are some thoughts I had about creativity, writing with speed, execution and rejection, and there are tons of other ways to get better at these things. Note that by this model, by far the most important quality in a person is the ability to reflect on their strengths and weaknesses and improve themselves in all of these categories.

I think this multiplicative model becomes even more interesting when you talk about working together with people in a lab. One point is that establishing a lab culture in which everyone pushes each other in all regards is critical and will have huge payoffs. Practically, this means having everyone buy in to what we collectively think of as a worthwhile idea, how we approach execution, how to write, what our standards of rigor are, and sharing stories of rejection and subsequent success through perseverance. This also provides some basis for the disastrous negative consequences of having a toxic person in lab: even if the effects on each other person in the lab in all or even some of these qualities are small, in aggregate, it can have a huge effect.

The other point is delegation strategy. It’s clear that in this model, one must avoid bottlenecks at all costs. This means that if you are unable to do something for reasons of time or otherwise and the person you are working with is also unable to do that task, things are going to get ugly. The most obvious case is that most PIs have only a limited capacity (if any) to actually work on a project. So if a trainee is unable to work on the project, nothing will happen. Next most obvious case is inability to write. If the trainee is unable to write and you as a PI have no time or desire to write, papers will not get written, period. Deciding how much time to invest in developing a trainee’s skills to shore up particular weaknesses is a related but somewhat different matter, and one that I think depends on the context.

This model also maybe provides some basis for the importance of “grit” or resilience or motor or drive or whatever it is you want to call it. These underlie those items on the list that are the hardest to change through mentorship. If someone just doesn’t have an ability to work on a project, then there’s not a whole lot you can do about it. If someone does not have the determination to do all the little things required to finish a project or to stick to it in the face of rejection, it will be hard to make progress, and there’s not much that you can do to alleviate these deficiencies as a mentor. I think many PIs have made this realization, and I have often gotten the advice that the most important thing they look for in a person is enthusiasm and drive. I would add to this being open to reflection and self-improvement. Everything else is just gravy.

No comments:

Post a Comment