At first glance, luck seems to be clearly associated with psi. Both are related to the experience of favorable or meaningful events under slightly mysterious circumstances. We conducted a study to explore the relationship between attitudes and beliefs about luck versus performance on a psi task.

Previous psi research on luck

Experimental research on possible relationships between psi and luck began nearly six decades ago. Warner (1937) examined runs of luck using himself as the sole subject in an ESP experiment. Warner was partially motivated to investigate the notion of luck as a response of some of J.B. Rhine's critics, who suggested that Rhine used subjects in ESP tasks until they began to perform at chance levels. Warner wondered:

Can it be that investigators of telepathy and clairvoyance have discarded all subjects except those who, at the time, were having a long streak of luck? (Warner, 1937, p.85)

Warner conducted a clairvoyance test with Zener cards and hoped to detect runs of success which strongly deviated from expectation. His results were not significant.

Greene (1960) conducted an experiment looking at possible correlates between luck and psychokinesis (PK), and her questionnaire on luck was reused in subsequent tests (Ratte & Greene, 1960, and Ratte, 1960). It should be noted that the terms "luck" and "lucky" were only vaguely described in her paper as a "feeling of luck" and a "feeling of being lucky." That is, no definition was given for luck and it was assumed that everyone responding to her questionnaire had the same understanding of its meaning.

The experimental part of Greene's study was a test linking one's belief in luck with PK performance. The study involved a population of 281 students at Duke University. They completed a survey on luck and belief toward psi. The population was then reduced to 24 students who were judged to be outstandingly lucky or unlucky according to the results of their questionnaire. This group was further separated into two sets - those who reported success in winning at dice games and those who reported losing at dice games. The entire group then participated in a test of luck involving the tossing of dice and attempting to influence the outcome of the roll. Money was used as motivation and the subjects were told they could take home whatever money they won. The results showed that the self-reported unlucky persons performed slightly, but non-significantly, better than the lucky ones.

The Greene Luck Questionnaire resurfaced thirty years later in an experiment conducted by Gissuararson & Morris (1991). The luck questionnaire was used in one of five studies in their experimental series, and resulted in a non-significant negative correlation with PK performance.

Wiseman, Harris & Middleton (1994) then explored luck from a perceptual (ESP) as opposed to an active (PK) approach. Their experiment involved three different surveys: A questionnaire on luck, a questionnaire which asked the subject to report the likelihood of an event happening to them as opposed to another person, and a questionnaire which asked the extent to which the questions stated on the second questionnaire may have been due to chance. The luck questionnaire defined luck as follows:

Many events have their outcomes determined, at least in part, by chance. For the purposes of this questionnaire, if, over time, such events have seemed predominantly to work out well for you, you should consider yourself lucky. If, over time, such events have seemed predominantly to work out badly for you, you should consider yourself to be unlucky.

After completing all questionnaires, the subjects (N=57) were taken through a brief relaxation exercise and were asked to think about a target picture sealed inside an envelope. They were given five minutes to either write a description or draw their impression of the target. The experimenter then displayed a slid of four possible target pictures, one of which they had in their envelope. Subjects ranked each of the four pictures, according to their earlier impressions. Then they were asked to open their envelopes.

The experiment resulted in a significant positive correlation for subjects who believed that an ESP task was dependent on non-chance factors. For these subjects, ESP performance was related to their assessment of whether they thought they were currently lucky.