Academic journal article The Journal of Parapsychology

Reliabilities and Psychological Correlates of Guessing and Scoring Behavior in a Forced-Choice ESP Task

Academic journal article The Journal of Parapsychology

Reliabilities and Psychological Correlates of Guessing and Scoring Behavior in a Forced-Choice ESP Task

Article excerpt

A series of ten experiments designed to test the relationship between performance on a forced-choice ESP task and the Defense Mechanism Test (DMT) was carried out in Iceland between 1977-1991. These experiments confirmed the DMT-ESP relationship and, moreover, revealed a significant positive correlation of ESP performance with religiosity, and a significant negative correlation with psychotism of the subjects. The multivariate nature of these data allows various aspects of the data to be studied further (Haraldsson & Houtkooper, 1992).

A detailed analysis of defense mechanisms as rendered by the specific codings of the DMT has been carried out in an attempt to find a more objective and reliable way of rating the DMT. This attempt has delivered a linear combination of coded features predicting quite reliably the somewhat subjective DMT ratings by the human expert, but the predictor thus found does not correlate significantly with ESP performance (Houtkooper & Haraldsson, 1995).

Recently, the scoring patterns within the ESP task were investigated in an attempt to improve predictability of ESP scoring patterns by the DMT. It led to the conclusion that the best prediction is still the prediction of the overall ESP score by the DMT rating (Haraldsson & Houtkooper, 1995).

Furthermore, part of the DMT-ESP data were used in a study of the reliability of ESP scores and of the sheep-goat variable (the degree of belief in ESP) (Haraldsson, 1983). In this study, 98 subjects who in 1977-1979 had participated in three experiments, I to III, were retested on both variables in 1980. In these experiments, the ESP tasks had consisted of a 40-trial ESP test on a commercially available ESP-tester and 40 trials of a paper-and-pencil precognition test. The retest was carried out using a five-alternative test, implemented as a computer program on an Apple II computer with a built-in Zener diode noise-based random number generator. This investigation revealed a nonsignificant but positive (Pearson's r = .13) test-retest correlation of ESP scores and quite reliable (r = .78) sheep-goat scores. Reliability of ESP scores is generally low and not often reported (Palmer, 1977; Haraldsson, 1983).

Reliability is usually mentioned, as we do here, as equivalent to stability of scores. Theoretically, this involves lumping together the measurement error with the possible but unknown influences on the true level of the underlying variable and its fluctuation, for instance, from test to retest.

Thus, it seemed that the Icelandic DMT-ESP experimental data had delivered all there is to learn from it. Recently, however, attention has been drawn in the psychological literature to the purported relationship between belief in ESP (the sheep-goat variable, hereafter called psi-belie]) and guessing behavior, particularly "repetition avoidance," which is the tendency to call the last previously called target alternative (or symbol, in Zener card terms) again less often than a purely random guesser would (Brugger, Landis, & Regard, 1990). The possibility that ESP-believers avoid repetitions more often than non-believers do has been investigated by other researchers (Blackmore & Troscianko, 1985; Broughton, 1991) with little or no confirmation of the result of Brugger, et al. An earlier study of forced-choice ESP by Schouten (1973) focused on guessing behavior deviations from the purely random sequence. Schouten even trained some of his subjects to avoid these deviations, particularly bias, repetition avoidance, and "short term balancing." The latter is the tendency to guess all different target alternatives in a relatively short sequence. Schouten's persistent efforts did not achieve significant above-chance scoring.

In contrast to this, Palmer, analysing the data of a precognition experiment by Schmidt with a few high-scoring subjects, found strong non-random guessing behavior that was apparently not detrimental to ESP performance (Palmer, 1995). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.