Organizational Behavior and Human Decision Processes, Volume 70, Issue 3

Doctors modify drug dosage to bring disease states unSubjects playing the role of psychiatrists (actors) der control with minimal side effects. Stocktakers alter engaged in a simulated medical decision-making task the size of regular orders to ensure that demand is met in which they attempted to bring the value of a patient without excessive storage costs being incurred. Indusindicator variable into a desired range. For each treattrial process controllers regulate features of some onment recommended by the actor, both the actor and going reaction to ensure that an appropriate amount an observer subject playing the role of a nurse assessed of the desired material is produced within safety conthe probability that the treatment would be effective. straints. Recently, the processes underlying judgmental Both actors and observers were overconfident. Actors were more confident in their treatment recommendacontrol in tasks such as these have been studied by tions than were observers, but this difference was elimexamining performance with simulated systems (e.g., inated when observers were given the opportunity to Harvey, 1990a; Kleinmuntz & Thomas, 1987; Moray, offer their own alternative recommendation. Under Lootstein & Pajak, 1986; Sterman, 1989). the latter circumstances, actors and observers were In situations like the ones just outlined, the probabilequally confident in the actors’ decisions but observers ity that a proposed control decision will be effective were more confident than actors in the observers’ decioften has to be estimated. This is because the executive sions. These findings suggest that while control over decision of whether to implement it must take into acthe outcome of the decision has little influence on count both the expense of changing control parameters actor–observer differences in confidence, feedback regarding this outcome plays a crucial role. q 1997 and the possibility that the costs of moving further