Poster F82, Tuesday, March 27, 8:00-10:00 am, Exhibit Hall C
Reverse Inference Problem with Task Difficulty and Reaction Times
Alexander Gordon1, Mark Straccia1, Matthew Lieberman1; 1University of California, Los Angeles
Perhaps the most widely used and robust measure of task difficulty is response time and error rate. However, while a more difficult task generally takes longer and has more errors, observing these effects does not guarantee the task is more difficult– a reverse inference claim (since other effects might be the cause). To support this view, we use another robust indicator of task difficulty, increased pupil dilation, to assess (with an eye-tracker) processing load put on participants. We used a paradigm by Gilbert et al. (2012) where participants (n=35) either a) respond if a letter has a curve b) hold in working memory the next letter in a sequence and respond whether it has a curve or c) respond if a novel line drawing has a curve. We replicate Gilbert et al.’s (2012) finding that the novel line drawing trials have the longest reactions times and most errors; however, we also show that participants had the least pupil dilation responding to the novel line drawing and the most pupil dilation to the working memory letter sequence demonstrating more cognitive effort for this task. The pupil dilation pattern observed here suggests that the medial prefrontal pattern observed in Gilbert et al. was due to cognitive demand, in contrast to their account. More importantly our results demonstrate experimenters need to be more aware other factors might cause longer reaction times (e.g., orientation to a stimulus, visual search time, etc.) rather than task difficulty to avoid the reverse inference mistake.
Topic Area: METHODS: Other