Response randomization of one- and two-person rock-paper-scissors games in individuals with schizophrenia

Abstract

Randomization among successive choices is important in adaptive decision-making, particularly for strategic interactions in which the optimal strategy is a mixed strategy. Patients with schizophrenia have been reported to have deficits in random sequential behaviors arising from impaired executive function. However, whether schizophrenic patients exhibit distinct behaviors for response randomization in one- and two-person games requiring different behavioral strategies is not known. The aim of this study was to examine the response randomization of 48 schizophrenic patients and 50 healthy subjects in one- and two-person rock-paper-scissors games. Here we found that the schizophrenic patients exhibited non-random biases distinct from those of the healthy subjects (i.e., stereotypic switching in the one-person game and the tendency to choose the best response against the opponent's previous choice in the two-person game). The entropy of the choice sequences was prominently decreased in the schizophrenic patients for both games, thereby indicating an overall disturbance in the behavioral randomization in adaptive decision-making. These results suggest that the impairment of response randomization in schizophrenic patients manifests differently in interactive and non-interactive situations, which may be useful for the diagnosis and quantification of the severity of the disease.