I run a mid sized cognitive psychology lab: it’s me as the PI, 2 PhD students 3 master’s students and a handful of undergraduate honours students and RAs. We are a reasonably productive lab, but there are times when I think we could be doing more in terms of getting our work out and also coming up with innovative and creative ideas.
Lately I’ve been thinking of ways to break out of our routines. Research, in my opinion, should be a combination of repetition (writing, collecting data, running an analysis in R) but also innovation where we look at new techniques, new ideas, new explanations. How to balance these? Also, I want to increase collaborative problem solving in my lab. Often a student has a data set and the most common process is the student and I working together, or me reviewing what she or he has done. But sometimes it would be great if we’re all aware of the challenges and promises of each other’s work. We have weekly lab meetings, but that’s not always enough.
What follows are some ideas I’d like to implement in the near future. I’d love to hear what works (and does not work) from other scientists.
An Afternoon of Code
We rely on software (PsychoPy, Python, R, and E-Prime) to collect behavioural data. We have several decent programs to run the experiments we want to run, but that is often a bottleneck, and all of us sometime struggle to translate ideas into code. One way to work on this might be to have a coding retreat or an afternoon of coding. We all agree to meet in my lab and we work on shared task or designing a paradigm that we’ve never used before. I’d put up a prize for the first student to solve the problem. As an example, I’m looking to get a version of the classic “weather prediction task“. We might agree to spend a day working on this, maybe each on our own program, but at the same time so we can share ideas.
Data Visualization and Analysis
Similar to the idea above, I am thinking of ways to improve our skills on R-Studio. One idea might be to have a set of data from the most recent study in our lab and we spend a day working together on R-Studio to explore different visualizations, techniques for parsing, etc. We each know different things and R allows for so much customization, but it would be helpful to be aware of each other’s skill set.
Writing at the Pub
Despite some of its limitations, I’ve been using Google Docs as a way to prepare manuscripts for publication. It’s not much worse than Word but really allows for better collaborative work and integrates smoothly with #Slack. With the addition of Paperpile, it’s a very competent document preparation system. So I thought about setting aside a few hours in the campus pub, bring our laptops, and all write together. Lab members that are working together on a paper can write simultaneously. Or we might pick one paper, and even grad students who are not authors per se would still be able to help with edits and idea. Maybe start with coffee/tea…then a beer or two.
I’ve also thought about spending some time designing and implementing replications of earlier work. We already do this to some degree, but I have many published studies from 10 or more years ago that might be worth revisiting. I thought of meeting once every few month with my team to look at these and pick one to replicate. Then we work as a team to try to replicate the study as if it were someone else’s work (not ours) and run a full study. This would be done along side the new/current work in our lab.
Chefs learn by repeating the basic techniques over and over again until they master them and can produce a simple dish perfectly each time. I can think of no reason not to employ the same technique in my lab. I think the repetitive, inward focused nature of a task like this might also lead to new insights as we rediscover what led up to design a task or experiment in a certain way.
I am planning on taking these ideas to my trainees at a one of our weekly lab in the next few weeks. My goal is to just try a few new things to break up the routine. I’d welcome any comments, ideas, or suggestions.