This week I’ve been running an experiment I first ran in 2006, collecting more data to see if some of the effects would actually reach statistical significance with more data, or would the inklings of trends be shown to be just flukes.
I’ll write more about the results of this experiment later, I just wanted to share some of my first impressions. These first impressions are mainly based on post-experiment conversations with participants, and in my opinion are important not only when interpreting the results, but also when trying to tie the experiment into its wider contexts. I feel that I learn at least as much about the phenomenon by talking to people after they’ve performed my experiment as I learn by looking through their data. Data is needed to find evidence for or against a hypothesis, but learning about the stuff you’re working with is a more all-encompassing process.