Mind reading: discovering individual preferences from eye movements using switching hidden Markov models

Abstract

Here we used a hidden Markov model (HMM) based approach to infer individual choices from eye movements in preference decision-making. We assumed that during a decision making process, participants may switch between exploration and decision-making periods, and this behavior can be better captured with a Switching HMM (SHMM). Through clustering individual eye movement patterns described in SHMMs, we automatically discovered two groups of participants with different decision making behavior. One group showed a strong and early bias to look more often at the to-be chosen stimulus (i.e., the gaze cascade effect; Shimojo et al., 2003) with a short final decision-making period. The other group showed a weaker cascade effect with a longer final decision- making period. The SHMMs also showed capable of inferring participants’ preference choice on each trial with high accuracy. Thus, our SHMM approach made it possible to reveal individual differences in decision making and discover individual preferences from eye movement data.