Gender Effects on “Predictions” of Computer-Random Numbers as a Function of Feedback Conditions

Abstract

If “randomlike” behavior can be learned in a laboratory setting, a question is the extent to which men and women can profit from systematic feedback in successfully predicting random events. Undergraduate volunteers (34 men and 78 women) were asked to predict the identity of computer-random numbers arranged in sets of 25 digits (1 to 5) and displayed individually on a Televideo computer terminal. Men and women (between-subjects) were instructed to predict 4 sets of 25 random digits, and all subjects were given immediate feedback of the correct target for 2 of the four sets (within-subjects) in a counterbalanced design. Chi-squared tests of the distribution of predictions showed significant gender dependencies. Also, the distribution of predictions was significantly dependent upon whether the subjects received feedback after making a prediction. Implications for “psi” research and “chaos” theory are discussed.