Learning Regularities in a Sequence of Decision-Making Tasks

When people make decisions, these often do not stand alone but are integrated into a sequence of decisions. For instance, a doctor will first decide on a patient's treatment and then about the duration of the treatment. In such decision sequences, later decisions frequently depend on the outcome of the first decision. Grammar learning shows that humans can learn regularities between sequentially presented information. However, there is little to no research on the role of sequential regularities in more complex tasks such as category learning or judgment. Here, we investigate whether people can pick up regularities between the outcomes of two categorization tasks and use them to speed up learning and categorize novel items. In the experiment, we varied whether a contingency between the outcome of two categorization tasks existed and whether the two tasks were adjacent (the tasks followed each other) or non-adjacent (an estimation task took place between them). In the adjacent condition, participants learned to categorize the objects of the second task better than in the control and the non-adjacent condition. But, during transfer, participants used the dependency to categorize novel objects in the adjacent and the non-adjacent condition. These results are consistent with grammar learning experiments, indicating that humans can pick up regularities and learn adjacent ones more easily than non-adjacent ones. The results also show the importance of considering sequential regularities between decision tasks. unknown unknown