Exploring the Reliability of the Presentiment Effect

This project extends a series of experiments begun by Radin that appears to demonstrate an autonomic nervous system (ANS) response to future emotionally arousing experiences. This experiment used skin conductance as the ANS measure and was designed to examine test-retest reliability in the “presentiment” effect, also known as “prestimulus response” (PSR). The experiment also looked for individual differences in PSR responding using the Myers-Briggs Type Indicator and the NEO-FFI personality inventories. Volunteer participants completed two sessions of viewing 40 calm and arousing images from the International Affective Picture System (IAPS) while SC was monitored. The PSR window was defined as 3 seconds before stimulus onset, and the measure of PSR was the area between the mean relative SC levels for the two classes of stimuli within the prestimulus window. Significance was determined using a Monte Carlo method. A novel analysis technique of de-trending the SC data immediately before the PSR analysis window was used to counter the possible effect of expectancy and other artifacts. Eighty planned participants were tested. Sixteen were eliminated from further analysis because of lack of any SCRs, leaving 64 participants in the final data pool. The experiment did not produce overall evidence of a presentiment effect, nor did it demonstrate test-retest reliability. Since this experiment deviated from prior methods of analysis the data were examined without detrending and automatic artifact removal, but there was still no evidence of a presentiment effect. Of three personality factors previously associated with better ESP scoring, MBTI Extraversion, MBTI Intuition, and NEO Openness, the last two were positively and significantly correlated with individual mean PSR. The data were explicitly tested for the presence of an expectancy artifact by correlating the PSR prior to arousing stimuli with the time since the last arousing stimulus. The correlations for both the de-trended data and the original method of analysis were negligible and non-significant, providing no evidence that an expectancy artifact was present.