January 21, 2010

Building on success

One of the other things I've been musing about since the Big National Conference is just how much feedback there is in My Science: in order for people to take you seriously, they have to know your name already, and once you're Somebody it is much easier to get the research funding and data access you need to do more good work. I don't assume that people have heard of me, with my mighty five (I think) publications in my old-school subfield, but I am trying to raise my profile by moving into sexier adjacent subfields, and by going to meetings and talking about my projects to try to make an in-person impression.

I have a good friend who is already On His Way. E is tremendously dedicated and creative, and also lucky to have thought up a timely and clever project with a well-known and socially adept advisor. He's a good speaker, the right people think well of him, and he just landed a highly respected research job. I'm glad he caught that wave, and hope he stays on it: he'll be great.

We have conceptual feedback as well, of course - research scientists tend to think of ourselves as meritocrats, respecting well-done and convincing work, and if Theory A was well-done and convincing, why should we entertain the possibility of Theory B? There's some ego acting here too: we couldn't have believed the wrong explanation for so long!

These are both deeply human tendencies, and will always play some role in the human scientific community. They can be mitigated with attention, and that's probably good: this article was eye-opening for me (in that makes-you-want-to-hide-under-the-bed kind of way). The effort to be fairer is both individual and institutional (or at least that's how I was thinking about it yesterday), and I think that the effort to be more flexible is mainly cultural. It's our job to not know all the answers, to keep thinking up questions we haven't solved yet, but it is hard to maintain a let's-see-what-happens attitude in a be-right-or-else environment.

On the advice of a total stranger on the Internet (what could go wrong?), I'm making up a new resolution for every month of the year rather than trying to do one thing all year. January's is to be a little easier on myself, to be less harsh a critic, and I see an opportunity for a corollary resolution here: I'm going to give more of a chance to papers whose authors I've never heard of, and I will try to see it as a sign of inquiry, not cluelessness, when people consider alternative explanations and make themselves persuadable.

January 20, 2010

Admission

I went to the big national (US) conference for people in My Science a couple of weeks ago. Most people will claim that they find this conference very unproductive, because it is so general, but it is an important venue for job interviews and maintaining personal connections. I actually got quite a bit of work done with some far-flung collaborators, made a point of spending time with people whose general good will I will need in the future, and had an all-around fun class-reunion-type experience.

The meeting itself was overwhelmingly large, so I did not go to most of the talks or read most of the posters. This more focused approach still left me dead on my feet at the end of the first day, and the talks and discussions I did attend kept me busy and engaged. The overall experience sparked a couple of post ideas for me, and this is one of them: promoting demographic diversity is very important in My Science, is totally dependent on the good-faith effort of people who are already in the club, and is pretty much doomed in an admissions and promotion system where we are overwhelmed with applicants and starved for time.

The last speaker in a session I went to on mentoring talked about his experiences as a black man from a small college going into a big-time graduate program: outright hostility, awkward social dynamics, undergraduate classes, and a lasting commitment to making that experience better for students who would come later. He's now a professor at Florida State, and had a lot of interesting and practical things to say about making graduate programs attractive and accessible to people with backgrounds like his: rural, few family expectations for college, no lab experience, not white, few demographic role models in the field, and less likely to maintain confidence in the face of setbacks (and the whole game in My Science is psychological: you need to realize that everyone's insecure, even the really clever people, and you can't listen to your self-doubt).

The things he does to make FSU a more diverse place are these:
  • First, let them in
Obvious! But not done, and also not done in multiples. Admitting one test-case diversity representative is not going to change the systematic things that make your program a hard place for minorities to succeed.

  • Second, allow for remediation without stigma
If you admit students who went to small or struggling undergraduate schools, they may not have covered as much ground as the students who went to world-class schools. Giving them the tools to catch up, without pretending like having gone to a small school makes them somehow less intelligent, will get you a well-prepared graduate class.

  • Third, make a comfortable department community (peer mentors can make a big difference)
I spent this session listening, and not talking, as I have never been a nonwhite person in a [My Science] department, and was very disheartened to hear how strong the xkcd effect is: the pressure to represent everyone Like You, and not to show strain, and never to admit weakness or ask for help. Without those worries, you can convince yourself that everyone needs help sometimes, and having to ask for it isn't an indictment of All Of You People, and - poof! - you get help, and things aren't so hard. Explicitly removing shame from this situation can make a huge difference - if your students trust that you're working in good faith.
  • Fourth, advertise at meetings of minority scientists and at HBCs
Personal connections are a big deal in My Science, and a personal encouragement from someone who is in the club that you should apply to a particular grad school, and that you do deserve a place there, can make a huge impression on young students.
  • Fifth, involve your undergrads
Give them all the research experience and hands-on understanding you can: experience is one of the shorthands graduate schools use in evaluating commitment level and interest.

When I was in grad school, the students had some long conversations on the merits and methods of increasing diversity in My Science. The business-related arguments (that a diversity of backgrounds brings a diversity of problem-solving approaches and perspectives) don't really work in our field: at least from the insider's perspective, there's not much cultural to what we do. The main point, we figured, was justice: why should people gain or lose the opportunity to do science based on what they look like or where they come from?

Being on the graduate admissions committee for a year, it quickly became clear to me that "where you come from" has a lot of effects on where you can go: my institution was very highly ranked as a graduate school for My Science, and as such could be pretty baldly snobby about who was admitted. It didn't matter how good your grades were at the University of Southern North Dakota at Hoople; we've never heard of them, we don't know the people who wrote your letters of recommendation, so we don't know whether you are prepared to come here and keep up with people who had "good preparation" (read: known quantity) at the right universities as undergraduates. In order to maintain our prestige, we have to admit mostly/only students who we are confident will do well.

This is pretty obviously problematic for people who want to go to grad school to learn things, or to become professional scientists, and not just to boost the status of a particular school by their effortless brilliance. Overall, and this is something I was shocked to realize I hadn't been using as an admissions standard, admissions ought to be about what students can do, not just what they have already done. It's very difficult to evaluate that, and when you have three days to rank 150 applications, it's so easy to start using shorthands (where an applicant went to college, GRE scores, name recognition of their letter-writers). If your department is willing to work on cultivating the students, and not just putting them to work in the data-reduction mines, I don't think it would be that big of a gamble to shake up the admissions process.