Everyone has heard the phrase “practice makes perfect”. As a kid, my dad had his own spin on that line: “If you’re going to practice garbage, you’re going to play like garbage”. I lettered all four years in high school playing golf, but never seemed to get much better over those four years. At the same time, I loved sitting on the driving range all day hitting ball after ball. Just hitting balls. I didn’t have much of a plan and I guess it showed. Fast forward 20 years, I’ve been in a tenure track position for ten years now and I wonder whether I’m getting any better. I’d like to think that I suck less than I did ten years ago, or at least suck differently than I did then, but how do I know? Am I improving in the areas that I really need to improve? When I resubmit grant proposals or resubmit a manuscript, am I merely responding to the previous rounds just enough to get over the hurdles or am I getting better?

I recently read “Peak: Secrets from the New Science of Expertise” by Anders Ericsson. In a way this is a rebuttal of “Outliers: The Story of Success” by Malcolm Gladwell. Really, the two books are quite complementary. I think the most important point of both books is that natural talent is overrated. People seem gifted, but they really bust their butts to get where they are. Chess grandmasters are grandmasters because they study the game. They practice. They learn from their mistakes. They have a coach. They don’t just sit around playing chess all day. As Ericsson expands from Gladwell’s “10,000 hour rule”, their practice is deliberate. Experts who don’t continue to practice will fall behind in their skills. I was left somewhat frustrated by “Peak”, because a lot of the applications seem a bit worthless to normal life: chess, perfect pitch, memorizing numbers, kids in the spelling bee, Scrabble, sports, etc. As an academic scientist who runs a lab, what does deliberate practice look like?

Perhaps you’re thinking that the distinction between practice and performance is irrelevant to an academic. Let’s say I get to the end of writing my latest grant proposal where I’ve meticulously followed the suggestions of a guide book and realize it kind of sucks. I said everything I needed to, but it’s not very inspiring. The important points get muddled and I went in a direction that really doesn’t fit with where I meant to go. Perhaps the analysis plan is pretty safe. It’s what we’ve always done and we’ll just plug and chug to get the work done. Practically speaking I should delete the file and start over. But I don’t have time for that. I’ve already invested more than two weeks writing the document and can’t start over. The deadline is in a few days. What do I do? I put lipstick on a pig. I think the same thing happens in running experiments, doing data analysis, and giving talks. We learn a technique from a workshop or protocol, but that doesn’t necessarily mean we’re any good at doing it or integrating it with other methods or inserting it into different circumstances. By using performance as a substitute for practice, we don’t make long-term strides in improvement. Ultimately, perhaps people get to be “dead wood” because they are continuing on their reserves from when they were a trainee and eventually just can’t keep up with where their own trainees have now taken the field.

Before reading on ask yourself, “How do I practice my craft as a scientist or as an academic?” I don’t mean how do you do science, but how do you get better at doing science? I also don’t mean how do you learn about a skill? Attending a workshop or reading a programming book doesn’t mean you’ve learned how to do a new method or to program, it just means you’ve been exposed to the material. I don’t have any great answers and am anxious to hear what others think. Let me throw out a few thoughts and I’ll look to see what people put in the comments…

  • Take on miniature writing projects. Create a goal for each piece of writing - coming up with a good lede, forming transitions, concluding the story, etc. Ask others to evaluate how you did on that aspect of the writing.
  • Write a bunch of specific aims pages for fictitious projects. Write a specific aims page for a project looking at the interaction between the microbiome and midi-chlorians. Go crazy. Share them with friends asking them for feedback on specific points.
  • For each talk you give, create a goal for yourself Perhaps it’s reducing text on slides, smiling, making eye contact, a new way to lead the audience through complex model, etc. Record yourself and evaluate whether you met the goal or ask others to explain a key point back to you.
  • Teach. If you really want to learn something, try to teach it. This will likely show you that you are pretty competent at what you do, it will teach you a few new things, and it will expose gaps in your knowledge. Ask for feedback from the learners as well as from colleagues.
  • Read. A lot. Learn from the science, but also from their writing. What do you like about each paper? Are there figures that you could see incorporating into your work? What would you do differently? The same could be done for each seminar you go to.
  • Review papers. Learn how others are presenting similar science. Crystalize your thinking on what type of data are needed to make a specific assertion. Most journals will supply you with the reviews from the other reviewers, which enables you to see how stringent or lenient you are relative to your peers.
  • Sit on study sections. What annoys me about other people’s science is usually something that annoys me about my own. Look for ideas on how people are testing hypotheses or dealing with third rails in your discipline. Also, when you sit with other scientists you learn what they are looking for when they’ll be reviewing your proposal.
  • Recreate figures. Try to recreate the latest plot from that paper you just read. If you don’t have the underlying data, make it up. Get someone else to make the figure in parallel to you and compare your approach. For a twist, do this activity with someone else working at the same computer with you where you take turns telling the other person what to do.
  • Present your code to others. You will learn a lot about programming methods and styles by putting your code in front of other people’s eye balls. They may know a more efficient way to do something, they will learn something new, and they will reveal points in your code that thought were clear but aren’t.
  • Do the questions at the end of the chapter. Frequently programming books will have problems at the end of the chapters. Ditto for statistics books. Stephen Heard’s new book on writing even has activities at the end of the chapters. Do them! Write the author and ask for the answers so you can check your work.
  • Spend time contributing to online forums. Go to stackoverflow or SEQAnswers and look for questions that you can answer. Read other people’s posts. You’ll get an evaluation from others on whether your answer was right, but you’ll also see how others address the same problem.
  • Get involved in an outside activity. Find an activity where you can test your skills doing something you find fun. Write a newsletter for the PTA, do data analysis for that local non-profit, volunteer to give a presentation for your organization in front of the planning board.

Ericsson is very clear that one of the reasons that it is hard to develop expertise later in life is that we just don’t have time. We have kids who have their own demands and some people have additional challenges that others don’t. I would like to think that each of these ideas could be part of our normal jobs. How much time should we take to practice our craft? I suspect it’s on the order of two hours a day. Just like our colleagues who train to run a marathon by following a detailed regimen over the course of months, we also have to come up with a practice plan. The hope is that by spending 20% of your day practicing, that the other 80% will be more efficient and just better. To be deliberate, our practice must be daily.