A comparison of temporal binding across different tasks
Temporal binding refers to the phenomenon through which a cause and its consequence appear closer in time than two unrelated events. Over the years, researchers have used several methods to investigate this effect, including methods that measure the ability to subjectively estimate a time interval (interval timing) and our ability to accurately say when an event occurred in time (occurrence timing). Recent evidence suggests that these two skills have distinct neural and cognitive bases, despite being complementary. Assuming that these differences must be reflected in the temporal binding effect, we implemented two experiments containing four tasks: two related to interval timing (Temporal Estimation and Reproduction) and two to occurrence timing (Temporal Anticipation and Libet Clock). Participants performed these tasks in two (Experiment 1) or six (Experiment 2) separate experimental sessions, and we correlated their data within and between sessions and tasks. Furthermore, because highly replicable effects suffer from a low between-subjects variance, limiting the correlational analysis, we estimated the reliabilities of the measures in each task and session. Across both experiments, we replicated the temporal binding effect in three tasks (Temporal Estimation, Reproduction, and Libet Clock) but not on Temporal Anticipation. We found good reliability of the effects within the same task and session but to a lesser degree between different experimental sessions. Lastly, we observed a correlation of binding effects across participants for Temporal Estimation and Reproduction but not for the other tasks. Taken together, our findings suggest that temporal binding is a robust experimental finding, its effect within the same participant varies across sessions and that different tasks can measure different aspects of the effect.