Social Influences in Behavioral Rigidity: Shared Attention in a Three-Alternative Choice Task

The understanding of the factors and conditions under which humans and other animals fail to adjust to changing circumstances is still incomplete. One possibility has to do with the role of social influences that might favor or hinder adequate control by contingencies. Three experiments are presented in which one such factor was manipulated, the presence of a coattending individual. Shared attention has been found to enhance cognitive resources devoted to a task in a variety of experimental settings. Participants played a computerized version of the Rock-Paper-Scissors game which was programmed so that a particular choice would be more likely to win than the remaining two within a block. The allocation of behavior towards these choices was used as a measure of the participants’ ability to track unannounced changes in the probabilities of the program’s choices and considered a measure of sensitivity. Although, shared attention had little effect on the overall outcomes, it had unsystematic effects on the participants’ strategy (Exp. 1). A significant effect of the sequence of block presentation was found when a progressively more discriminable sequence of blocks was compared to a progressively less discriminable one (Exp. 2). Early failure appeared to hinder effective control by the programmed probabilities. This effect is attributed to verbal behavior concurrent to performance. In support of this, providing participants with accurate instructions increased the overall accuracy, and these instructions were still followed to an extent when they became inaccurate (Exp. 3). Instructions had an immediate effect in the participant’s strategy, reflected in a reduced tendency to shift strategy following losses and, particularly, draws. Further manipulations of shared attention are required to ascribe a particular role of this social phenomenon to the rigidity of behavior and other instances of suboptimal choice