Effects of dyadic work organization on realism in confidence judgments

This study examines the effect of three regimens (organizations of activities) on the realism in dyad’s confidence judgments. The effect of letting the pair members first work alone when selecting an answer and/or when confidence rating the correctness of this answer was compared with working directly in pairs. No improvement in the realism of the dyads’ final confidence judgments was found as an effect of the work regimens involving preparatory work, compared with pairs having no such preparation. In fact, carrying out a confidence judgment acting as an individual after a dyadic decision about the selection of answer alternative, led to increased overconfidence in the pairs. Improved realism was found when pair members worked individually when selecting and confidence judging the answer, before collectively deciding on an answer but again giving individual confidence ratings, but this effect did not reach significance. It is suggested that these results might at least partly be explained by increased confidence as an effect of explaining one’s belief. Finally, each pair’s final judgment of the total number of questions answered correctly was not improved by individual preparatory work.