Mirror on Naive Protocol
Total Page:16
File Type:pdf, Size:1020Kb
unknown animal whose ancestry fails to exhibit a convincing pedigree and for whom hard evidence has failed to materialize even after years of diligent investi gation? The Scottish Tourist Board will not like this book. • Mirror on Naive Protocol Know Your Own Psi-Q. By Hans Eysenck and Carl Sargent. World Almanac Publications, New York, 1984. 176 pp. Reviewed by James Randi The American edition of Know Your Own Psi-Q, a U.K. original, is published by World Almanac Publications. In 1978, George E. Delury, then editor of The World Almanac, consulted CSICOP when preparing the Book of the Strange. The result (see Book Reviews, SI, Summer 1978) was a well-tempered account of a varied selection of claimed wonders. Were this gentleman still the editor, 1 doubt that publication of the U.S. edition of this Eysenck-Sargent effort would have been undertaken by World Almanac. The book is a disaster in every way except one: It may provide us with an accurate picture of just how naive the authors are in designing proper protocol for testing of psi-powers. If their book correctly expresses the standards of para- psychologists in general, it is no wonder that the rest of the scientific community scoffs at their efforts. In a section called "Psi Stars" the authors gush over the claimed miracles of D. D. Home, whose career—details of which are safely blurred by decades of elapsed time—is meant to inspire their readers with the wonders they hope to establish. Then they blantantly introduce modern "Psi-Star" Bill Delmore, saying, ". Delmore was successful with everything thrown at him. It's normal for researchers, given such a Psi-Star, to state that the odds against his score being due to chance are so many billions to one; but with Delmore, it's not worth bothering, for the word 'billion' would have to be repeated too many times." They conveniently forget to mention that prominent statistician (and accomplished card-conjuror) Persi Diaconis of Stanford University witnessed a session at Harvard in which Delmore performed informally some of the "controlled" tests that had earned him a reputation among the parapsychologists as well as a grant from Harvard University "to explore the nature of his own psychic ability." Diaconis found Delmore using the standard methods of card cheats and some obvious ploys quite familiar to conjurors, to deceive the observers. Yet Eysenck and Sargent enshrine him as a "Psi-Star." Another superstar they drag out of obscurity is the well-discredited Jean- Paul Girard, a French conjuror who stated long ago that he would use his 180 THE SKEPTICAL INQUIRER, Vol. 8 sleight-of-hand to demonstrate that dedicated believers among the scientists could be and would be fooled. They were—even after Girard failed a definitive set of tests (see S7, Winter 1978). But these authors still accept him as genuine, and so present him to their readers. Incredibly enough, these two Ph.D.'s were unable to see a rather basic statistical problem with a quite simple test they suggest in their book. It involves playing cards, and the subjects are required to guess only the color (red or black) of each card. Now the chances of doing this—by chance alone, without psi power—are ordinarily one in two, or 50 percent. Certain variances from this result become significant, and tables are provided by the authors to determine whether no, some, good, or excellent psi results are thus obtained. But in this test, the authors instruct the experimenter: "Tell the subject whether they are right or wrong." It is obvious even to the amateur observer that providing such feedback can increase the chance expectation. Each time the subject knows the color of a discard, information is available about the remaining cards. (In the case of the Zener ESP cards, where five of each of five different symbols are used, Diaconis tells us that if the subject is told what each card is after guessing it, the chance expectation jumps from 5 in 25 to 8.65 in 25!) There is an almost total lack of basic experimental caveats in the book. Even when one is provided—"Treat all the subjects the same way"—it is canceled out by the statement, "... unless any of them are vehement that they really don't want this circumstance." One of the most obvious pitfalls—optional stopping—is even condoned for use by the reader. "Try and fix the number of guessers to be tested in advance of the experiment, but if it really isn't possible to test as many as you would have liked, then stop when convenient." That "convenient" time might also be just when the guesser is ahead, too. The only proper treatment of aborted tests is to throw them out, but the authors remind us that this is not a "controlled laboratory study." In that case, it is a child's game that could have been properly designed to be much more than that. It is, unfortunately, rather typical of much "controlled" work in parapsychology—particularly some that involved one of these two authors. In Game No. 8, Know Your Own Psi-Q outlines a direct test (odds: one in two) that involves pictures of human males and females. Aside from the fact that there are no instructions about what the subject should be told concerning the possible ratio of male/female identities, this test seems fairly good. But instruc tion number three is revealing of the methodology that damns such tests when done in laboratories. It says: "The experimenter hands the envelopes one by one to the subject, who guesses the sex of the people in the pictures, taking as much time as necessary. The experimenter then writes down any impressions that the subject may have." (Italics added.) In a simple test, essentially a heads-or-tails matter, the authors have thrown in an extraneous factor that provides a tempting opportunity to extend, amplify, and save any mediocre results. Suppose a photo of Clark Gable is the target. The subject guesses "male" and goes on to say that he sees a mustache as well. How is that evaluated? Certainly it is a "hit" and is scored as such. But do we throw an extra point in for the mustache? We aren't told. And suppose that for the same target the guess is "female" and long, blond hair is included in the description. The guess is a "miss," but do we ignore the hair, or count it against the score? Such peripheral information has no place in a proper test, ever. Winter 1983-84 181 The "Computer Programs" section of the book consists of some interesting software treatments of psi-testing that if stripped of some of the juvenile scenar ios could be useful. Finally, in the bibliography, Eysenck and Sargent provide us with two and one-half pages of pro-paranormal journals, books, and other material, but not one skeptical work is made available to the reader. This clearly expresses the intent of the book, whether conscious or not: It is designed to give a totally one-sided view of the subject and to lead the reader to believe, as a result of faulty and biased testing procedures, that he himself has psychic powers. Given the tastes of the public, it will probably be quite successful in that respect. But I feel that the World Almanac publishers will be embarrassed by this book. • Some Recent Books Listing here does not preclude more detailed review in a future issue. MacDougall, Curtis D. Superstition and the Press. Prometheus Books, Buffalo, N.Y., 615 pp. $29.95, cloth; $16.95, paper. Everything you want to know about what the major daily newspapers have said about the para normal in recent years. An exhaustive, comprehensive compilation, by a noted journalism educator, of news coverage on twenty-seven paranormal and fringe-science subjects from astrology and artifacts to UFOs and witch craft. MacDougall's treatment is extensive rather than intensive—most summaries are only a paragraph or two. His style is matter of fact: there is a minimum of commentary and interpretation. The latter lack is at times annoying, but as MacDougall says, "My purpose is only to record what a [newspaper] reader thinks he knows." More often than not what the reader learned was painfully short of the full truth. (For some of MacDougall's comments while this work was in progress see the Summer 1982 SI, p. 9.) A valuable source book for charting media coverage and tracking the history of hundreds of claims and controversies about the paranormal. —Kendrick Frazier Articles of Note Baker, Robert A. "The Effect of Suggestion on Past-Lives Regression." American Journal of Clinical Hypnosis. 25(l):7l-76, July 1982. Controlled exper- 182 THE SKEPTICAL INQUIRER, Vol. 8 iment with three groups of subjects demonstrates that suggestions suppor tive of age regression prior to hypnosis are reflected in the "memories" recalled. "The present study offers additional evidence that past lives re gression phenomena rather than being examples of the 'reality' of reincar nation are, instead, the result of suggestions made by the hypnotist, ex pectations held by the subjects, and the demand characteristics of the hypnoidal relationship." Brandon, Ruth. "Prestidigitations" (letter). New Scientist, July 14, 1983, p. 139. Brief but strong response to a criticism by Brian Inglis of her article "Scientists and the Supernormal" (June 16). Clark, Jerome. "Scandal in the Spirit World." Fate, November 1983, pp. 64-67. Report on successful trap to catch British medium Paul McElhoney in act of fraud, and the reaction to the revelation.