EFFECTS OF CONSEQUENCES OF ADVICE ON PATTERNS OF RULE CONTROL AND RULE CHOICE

Rules in the form of advice can inaccurately state the effects of reco mmended responses by overstating or understating size of the consequen ces. Three experiments investigated the effects of such inaccuracies o n patterns of rule control and rule choice with female college student s. In Experiment 1, signaled accurate, overstated, or understated rule s specified that a given number of points would be earned by pressing a designated key. For some subjects, rules specified a number of point s to be gained; for other subjects, rules specified a number of points to be lost from an amount given earlier. Point totals stated in the i naccurate rules averaged 25% more (overstated) or 25% less (understate d) than those received. When subjects could choose either the response specified in the rule or an alternative response that produced an unp redictable number of points, they showed greater sensitivity to the in accuracy of overstated rules than understated rules. In trials at the end of the experiment in which subjects could choose which rule to see , subjects did not always choose accurate rules and often chose inaccu rate rules for which they had shown less sensitivity earlier. Experime nt 2 replicated this pattern in which subjects could choose which type of rule to see on a greater number of trials. Some evidence suggested that subjects prefer an improvement from the outcomes promised to tho se later received. In Experiment 3, rules misstated by averages of 25% and 50% were compared. Evidence suggested that increasing the size of the misstatement reduced the discrimination of inaccurate rules from accurate ones.