Psychic Staring Effect - Tests

Tests

In his controversial book The Sense Of Being Stared At, Sheldrake reviewed a 1913 study by E.Coover that concluded that although the feeling of being stared at is common, upon scientific examination it is revealed to be "groundless", as well as studies that reported some statistically relevant positive results, such as those by J. J. Poortman, Donald Petersen, Linda Williams and NEA Mao.

Sheldrake describes his experiments, one of which involved blindfolded subjects attempting to guess whether persons were staring at them or at another target. He says 60% of subjects reported being stared at when being stared at, and 50% of subjects reported being stared at when they were not being stared at. According to Sheldrake, this suggested a weak sense of being stared at but no sense of not being stared at. In another test, he and his peers sat behind an audience during a quiz show at the BBC and engaged in recorded periods of staring and not staring. Sheldrake reports that when a subsequent videotape was given to a third party to count the number of times people in the audience glanced backward, the number of backward glances during the staring times was significantly higher than during the non-staring time.

Sheldrake summarized his case in the Journal of Consciousness Studies, saying he found a hit rate of 53.1% with two subjects "nearly always right, scoring way above chance levels". Sheldrake has since designed an online staring test to find more people that have the same performance for formal experiments.

David Marks and John Colwell, writing in the Skeptical Inquirer (2000), criticized the experimental procedures Sheldrake had developed for tests designed to demonstrate the existence of the staring effect. Apart from the fact that Sheldrake had encouraged the involvement of lay members of the public in research of the effect, Marks and Colwell suggested that the sequences used in tests followed the same patterning that people who guess and gamble like to follow, saying that "these guessing patterns have relatively few long runs and many alternations" and that the "non-randomness of test sequences" could result in "implicit or explicit pattern learning when feedback is provided." According to Marks and Colwell, "When the patterns being guessed mirror naturally occurring guessing patterns, the results could go above or below chance levels even without feedback. Thus significant results could occur purely from non-random guessing." Non-randomization is one of seven flaws in parapsychological research identified by Marks.

Sheldrake addressed Marks and Colwells' criticisms in the Skeptical Inquirer (2000), pointing out that if subjects were implicitly learning patterns in the test sequences, the results in staring and non-staring trials should have been symmetrical. In addition, Sheldrake pointed out that in the trials where he used counterbalanced sequences, he did not provide feedback to the subjects, thereby making it impossible they could have learned any patterns in the pseudorandom sequence. Finally, Sheldrake pointed out that in 40% of the 20,000 trials they didn't use a pseudorandom sequence at all, but rather used a coin flip.

Michael Shermer wrote in Scientific American (2005) that there were a number of objections to Sheldrake's experiments on the sense of being stared at, reiterating Marks' and Colwell's points about non-randomization and the use of unsupervised laypeople, and adding confirmation bias and experimenter bias to the list of potential problems; he concluded that Sheldrake's claim was unfalsifiable.

Sheldrake responded to Shermer's article in a letter to Scientific American denying Shermer's claim that he (Sheldrake) was proposing a "universal life force" and stating that Shermer "misrepresented the evidence for the sense of being stared at." He wrote that Shermer "tried to give the impression that the case rested on unsupervised tests by people using the experimental protocol on my web site ( www.sheldrake.org ), but this is not true. My own summary of the evidence and the independent meta-analysis by Dean Radin did not include the data from these unsupervised tests, but relied instead on the results of many thousands of trials already published in peer-reviewed journals."