Participant Interviews

For a study I’m working on, I needed to recruit 128 participants and meet with them to generate data. While I would describe the overall experience to be a positive one — interacting with people is never a bad thing, really — there are certainly some issues I faced.

First of all, the experiment protocol and script really needs to be hammered down before starting sessions. This was really critical; I thought I had the protocol down pat. Turns out the opposite was true; I found myself adding more clarifying statements to my participants than I had realized I needed before starting the interviews. Additionally, I tweaked little things that I didn’t realize beforehand that not explicitly specifying would cause issues. A simple example — I didn’t notice that I hadn’t specified the amount of mental rotations I was asking my participants to perform. The whole packet seemed unnecessary so for completeness I ended up asking for one page. Point being that the devil really is in the details — these fine-grained, granular details make a big difference.

Secondly, patience is a virtue. A major one. When dealing with human subjects, it’s important to always be “on” — to always come across as eager, natural, and understanding. This is, of course, because you’re asking for their help — Hey! Come in here and do stuff for us! — but there’s also a larger reason why you should be this way … you want to engage your participants in the work they’re doing. Acting sour, tired, or being short with your participants tends to induce a mirroring effect wherein they mimic the the personality traits of the experimenter and that will bleed into the data they’re generating. Not good! An engaged participant will generate true and honest data, the type of data that leads to real results and (hopefully) that lofty height: statistical significance. Tired participants are less inclined to do so; they will follow the path of least resistance which allows them to leave early.

Finally, you need to leave time to do the other things you have to do in life. I pursued the data for my study aggressively, seeing ~128 people in a 2 week period (10 days). And then again the following weeks after (two-session structure). This left me often exhausted seeing people from 10AM to 10PM at night with no (planned) breaks inbetween. “Hard working”, maybe, but not good for mental health. In the future, I’m resolving to limit those hours. :/