Abstract

Four experiments involved a new procedure for testing psychokinesis in animals. A total of 34 aggressive tropical fish were given a 20-sec mirror presentation (a positive reinforcer for these animals) as a feedback stimulus for hits on a Schmidt random event generator. The performance of these Ss was compared with that of 34 nonaggressive fish (for which the mirror was not reinforcing) and with the hit rate of the random generator during randomness tests conducted without the presence of fish in the apparatus. In Exps I (pilot) and II (1st confirmation), 100 trials were run on each of 4 days, with an intertrial interval of 20 sec. In Exps III and IV (2nd and 3rd confirmations), Ss were tested on only 1 day. In Exps I and II, the aggressive Ss exhibited psi-hitting on Day 1 only. There was a decline effect over the 4 test days. The aggressive Ss of Exp IV exhibited psi-hitting on their sole day of testing. The aggressive Ss of Exp III exhibited chance performance. The performance of the nonaggressive Ss in each of the 4 experiments and the scores of the random generator on randomness trials never differed from chance expectation. Collapsing across experiments and combining all Day 1 scores (Day 1 procedures were identical for all) yielded significant psi-hitting for aggressive Ss, chance performance for the nonaggressive Ss, and a difference closely approaching significance between aggressive and nonaggressive Ss. 17 ref)