THE EFFECTS OF FEEDBACK AND PREDICTABILITY ON JUDGMENT
GOLDSBERRY, BETTY SANDERS
Doctor of Philosophy
Previous research has found that when subjects are given cognitive feedback, they reach higher levels of achievement than when they are given outcome feedback. It was hypothesized that this finding was due in part to the predictability of the task environment since outcome feedback is at a distinct disadvantage as a sole means of conveying such information. A study was conducted to compare response and outcome feedback under three conditions which varied in terms of the predictability between actual and optimum criteria. The design included a control group receiving no feedback at all, two response groups differing in precision of feedback information, and two outcome feedback groups differing on a quantity dimension. Task predictability conditions averaged across five learning blocks were high (r = .94), moderate (r = .87) and low low (r = .71). The study also attempted to clarify the definition of feedback and to equate the availability of task information in the various feedback conditions that were compared. The results, however, did not support the above hypothesis. The utility of outcome feedback was inferior to that of response feedback under all three predictability conditions tested. In fact, an interaction revealed that the effect of increased predictability raised rather than lowered the disparity between outcome and response feedback performance. Generally, a decline in task predictability accompanied a decline in performance measured in terms of achievement, hit-rate, knowledge, and control. The results also revealed that a control group that received no feedback at all performed as well as or better than those that received feedback when the availability of task information was equated. Moreover, eliminating the memory requirement inherent in the use of outcome feedback only worsened performance. Similarly, adding precision to the response feedback condition beyond the level of mere directional error information did not improve performance. The principal conclusions to be drawn from these findings are: (a) increasing predictability improves judgment performance, (b) providing outcome feedback is more detrimental to performance than providing response or no feedback when a valid task structure is available, and (c) increasing predictability does not reduce the disparity between the effectiveness of outcome and response feedback.