Conditioned anticipatory outcome searching in humans

Two experiments used eye tracking to examine visual searching for expected outcomes in humans during an associative-learning task. In both, participants learned to press keys on a keyboard to activate weapons to repel invading spaceships in the presence of predictive "sensors." In both experiments, eye tracking showed that participants came to direct their overt visual attention to the portions of the screen where the spaceship would arrive during the presentation of the sensor in a cue-specific manner. Participants also directed attention to the weapon that was used to repel the spaceship. The same results were observed regardless of whether participants were responding on the keyboard, or not (Experiment 2). Pupil dilations occurred to the appearance of the spaceship from the first trial and occurred to the predictive sensor stimuli on later trials in both experiments, suggesting that they might be conditioned responses. In Experiment 2 participants again looked for the expected outcome, but dilations to the predictive stimuli were shown to be an artifact of responding and not due to conditioning. The discussion involves the implications for investigating attention to predictive stimuli using eye trackers, roles of context in behavior, and the utility of outcome searching and pupil dilations as indexes of learning.