Citation

Several psi and nonpsi hypotheses were tested in a computer guessing task. Participants (Ps) were 64 volunteers, 32 self-described strong believers in the paranormal and 32 self-described strong nonbelievers. Ps guessed sequences of the numbers 1–4 by calling each guess out loud and simultaneously clicking the mouse to register the response. In the 1st 2 runs, the target sequence reflected either pure repetition avoidance or pure counting, e.g., 2,3,4,1,2,3,4,1,2…. After Run 2, Ps completed psychological tests while the experimenter calculated their response bias in the preceding 2 runs. The 100 scored targets for Run 3 were random, except that every time P clicked the mouse when a computer address registered a 1 (1- state), which occurred randomly 20% of the time, they would receive a target for the next trial that matched their response bias in the preceding 2 runs, increasing the chances of a hit. As predicted from May’s decision augmentation theory, in Run 3 believers clicked the mouse significantly more frequently than chance when the computer was in the 1-state, and significantly more often than nonbelievers. Both random and total hits in Run 3 were positively and significantly correlated with scores on the AT-20 test of tolerance for ambiguity.