Children and adults distinguish irrelevant from relevant actions in over-imitation tasks – but some irrelevant actions are easier to identify than others
With this dataset, we investigated the ability of 4-year-old children and adults to discriminate between relevant and different types of irrelevant actions (pseudo-instrumental actions, noncontact actions, see Schleihauf & Hoehl, 2020). We applied a visual paired preference eye-tracking paradigm in a within-subject design. In each trial, we presented two video clips simultaneously showing actions of different action types. These actions were either functionally irrelevant or relevant for reaching a prior specified goal, i.e. extracting a golden marble from a puzzle box. Participants saw either (a) two nonidentical actions of the same action type (two irrelevant noncontact actions, two irrelevant pseudo-instrumental actions, or two relevant actions), (b) a relevant action paired with an irrelevant pseudo-instrumental action, (c) a relevant action paired with an irrelevant noncontact irrelevant action, or (d) an irrelevant pseudo-instrumental action paired with an irrelevant noncontact irrelevant action. Participants were instructed that after watching the videos they would be asked to extract a marble themselves. Therefore, we expected participants to direct their attention to the action which seemed most likely to be task-relevant. We measured their looking times to both simultaneously demonstrated actions, which allowed us to calculate a preference score that indicated whether they looked longer at one or the other action (for more details see the word document "Variable description"). We hypothesized that looking preferences are stronger when it is easy for participants to identify the task-relevant action, whereas we thought looking preferences are lower when it is harder for participants to identify the relevant action (for more details see the word document "Hypotheses").