Kevin A. Smith

Sources of Uncertainty in Intuitive Physics

Recent work suggests that people predict how objects interact in a manner consistent with Newtonian physics, but with additional uncertainty. However, the sources of uncertainty have not been examined. In this study, we measure perceptual noise in initial conditions and stochasticity in the physical model used to make predictions. Participants predicted the trajectory of a moving object through occluded motion and bounces, and we compared their behavior to an ideal observer model. We found that human judgments cannot be captured by simple heuristics and must incorporate noisy dynamics. Moreover, these judgments are biased consistently with a prior expectation on object destinations, suggesting that people use simple expectations about outcomes to compensate for uncertainty about their physical models.

Multiply-constrained semantic search in the Remote Associates Test.

Many important problems require consideration of multiple constraints, such as choosing a job based on salary, location, and responsibilities. We used the Remote Associates Test to study how people solve such multiply-constrained problems by asking participants to make guesses as they came to mind. We evaluated how people generated these guesses by using Latent Semantic Analysis to measure the similarity between the guesses, cues, and answers. We found that people use two systematic strategies to solve multiply-constrained problems: (a) people produce guesses primarily on the basis of just one of the three cues at a time; and (b) people adopt a local search strategy--they make new guesses based in part on their previous guesses. These results inform how people combine constraints to search through and retrieve semantic information from memory.

Similarities and differences in forward and reverse motion extrapolation

We often must not only extrapolate the future trajectory of objects (where will a thrown rock go?) but also extrapolate backwards in time (where did the rock flying by your head come from?). We matched forward and reverse extrapolation in two experiments to investigate similarities and differences between extrapolation directions. Forward and reverse extrapolation share biases and response time patterns across various trajectories, but reverse extrapolation is noisier. This suggests both forms of extrapolation share cognitive processes, and opens up further avenues of investigation into why reverse extrapolation is noisier. In Experiment 1, participants observed a ball moving either towards (forward) or away from (reverse) a semi-circular occluder, then a mark appeared on the outside of the semi-circle and participants indicated whether the ball would travel (forward) or came from (reverse) above or below this mark. 85% accuracy was maintained by dynamically adjusting the mark for each condition. Participants were slightly slower to respond when the occluder was larger (F(2,32)=4.8, p=0.015), but speed was unaffected by extrapolation type (F(1,16)=0.4, p=0.56). Reverse extrapolation was harder (greater offset thresholds: F(1,16)=5.4, p=0.034), though difficulty increased at the same rate over distance (no interaction: F(2,32)=0.04, p=0.96). These results suggest shared processing underlies both extrapolation directions, but do not differentiate whether reverse extrapolation is more biased or simply more variable. In Experiment 2, participants observed a ball in motion and indicated where on a line it would next cross (forward) or last came from (reverse). Each forward trial had a matched reverse trial with mirrored motion, so the line crossing would be identical. Systematic biases were nearly identical across extrapolation directions (r=0.96) but people were on average 30% more variable on reverse trials. This suggests that reverse extrapolation and forward extrapolation share biases, and that greater variability caused the increased difficulty in Experiment 1. Meeting abstract presented at VSS 2015.