Arnob, Entry #2 - Piloting my experiment

Last Friday my PI Mike was talking to Yul (the graduate student I am working under) and me about how far I was progressing with my experiment. After the pleasant talk, Mike said to me, "We are all going for drinks tonight. You can come if you want. You can drink soda or something." At this point I felt really awkward because I was caught off guard, and I ended up quickly, but respectfully declining his offer. This was a really nice gesture from Mike, and stuff like this shows that my lab members are really nice people. They can be incredibly hard-working on their arduous tasks in the lab, but they would also not hesitate to pause their work if I asked them a question on my own experiment.

Jian, another lab member, asked me to help him with his experiment. His decision-making experiment is similar to mine, where the naive subject chooses if the general coherence of a set of dots is moving left or right. But my experiment involves a track pad and pen that is connected to a computer, but Jian made an iPad app for his experiment. I am helping him by giving him feedback on his experiment, and I spend about thirty minutes everyday testing his app.

Jian's iPad app: the subject has to follow the yellow dot with his finger but cannot look at the yellow dot. The subject has to focus on the white dots above and judging by the coherence of those dots, the subject will choose whether to move his finger left or right.

Today I finally finished learning all the MATLAB code that I would need to properly analyze the data of the results of my experiment. I did a test run on myself and analyzed the data. One of the results of my analysis was a graph that showed when I had a change of mind and I changed my initial decision of left to right or vice versa. Figuring out how to make these graphs on MATLAB took me multiple days because the code was not simple, but the results look good.

This is a map of the movement of my hand during my experiment. The red lines show which in which trials I had a change of mind.

The overall goal of my experiment is to maximize the amount of changes of minds happen per session. To do this I needed to learn how to change the parameters of the experiment (also MATLAB based). There are other factors that I need to take into account, such as the machine that tracks eye movement to make sure the subject's eyes are looking at the right place. But I am looking forward to officially starting to pilot this experiment, which I will probably do after finishing this blog post!