Ren Calabro¹, Meriel Doyle¹, Kannon Bhattacharyya¹, Yuan Chang Leong¹; ¹University of Chicago

When reasoning about the world around them, people rely extensively on their inferences about physical scenarios. Examples of how these inferences can be flawed are well-documented, including in real-world optical illusions. In the current work, we test the hypothesis that our intuitive physical judgments can change when we are motivated to see a particular outcome in a physical scenario. While undergoing eye-tracking, participants (N = 57) were presented with images of block towers that varied in stability and were rewarded for correctly judging whether each tower would fall or remain standing under the influence of gravity. We also incentivized participants (N = 29) to judge towers as stable or unstable using monetary bonuses; the bonuses did not depend on the participants’ responses, so the reward-maximizing strategy was to respond accurately. We hypothesized that people would be motivated to respond that a given tower was stable versus unstable when financially incentivized to make a certain stability judgment. We found that on average, participants were more likely to judge the tower as stable when motivated to see it as stable than when motivated to see it as unstable. Moreover, we found that subjects’ fixation patterns were significantly correlated with the average fixation patterns of other subjects who made the same choice, suggesting that how people sample information contributes to their judgments about physical scenes. Participants’ fixation patterns were also significantly correlated with the average fixation patterns of other subjects who had the same motivation, indicating that motivation may influence how people sample information when making judgments. Further analysis will assess whether participants focus on a tower’s mechanical failure points and/or look where they predict the blocks may fall when they judge a given tower to be unstable. This study may shed light on potential limitations of human physical scene understanding.