The effects of unidirectional versus bidirectional rating procedures on college students' judgments of response-outcome contingency

Studies of human contingency judgment often reveal substantial bias and inaccuracy in ratings of cause-effect relationship. This study examined the effects of different rating procedures on contingency judgment. In two experiments, college students responded to and rated a range of positive, negative, and zero contingencies presented in a free-operant format. Experiment 1 found that judgments made along a unidirectional 0 to 100 “control” scale were less sensitive to variations in contingency (especially negative ones) and were based on less sophisticated judgment strategies than judgments made along a bidirectional − 100 to + 100 “prevent-cause” scale. Experiment 2 found that informing subjects given the unidirectional scale that their responding may either cause or prevent the outcome eliminated the tendency to underestimate negative contingencies and improved the sophistication of strategy use but did not increase judgment sensitivity to the level of subjects given the bidirectional scale. Taken together, these results suggest that the bidirectional “prevent-cause” scale contributes to highly accurate and unbiased ratings of response-outcome contingency.