Printer Friendly


DECEPTION AND SELF-DECEPTION: INVESTIGATING PSYCHICS by Richard Wiseman. Amherst, NY: Prometheus Books, 1997. Pp. 0 + 266. $25.95 (hardcover). ISBN 1-57392-121-1 [1]

This book is largely a collection of previously published journal papers, some from parapsychology journals and some from psychology journals. After an introduction for this volume, the chapters, collectively, consist of one review, one rudimentary conceptual model, one critique of a research report, and eight research reports. In the latter, Wiseman always is joined by one or more coauthors.

This volume's introduction describes Wiseman's personal history of interest in investigating claims of the paranormal, and it makes a case, using some dramatic, even terrifying, examples, that what soi-disant psychics offer the public is not always something benign. The latter circumstance is seen as indicating a need for careful scientific investigation of such claims.

The core of the volume begins ("Toward a Psychology of Deception") with a once-over-all-too-lightly review of the psychology of deception in a variety of settings that range far beyond the domain of the ostensibly psychic. This chapter touches on conjuring, psychic fraud, lying, confidence games, military deception, and animal deception. Although this chapter probably will tantalize more than it satisfies, its references will make useful reading for the seriously curious. Research citations, although plentiful and often very useful, nonetheless sometimes are not up to date, and the reader will miss some valuable, more contemporary, discussions of relevant topics than typically are cited in this volume. This problem is fairly common in other chapters, too.

Chapter 2, "Modeling the Stratagems of Psychic Fraud," with Robert L. Morris as junior author (material first presented in the European Journal of Parapsychology, 1994), is an effective treatment of the principles of pseudopsychic practice. The chapter seems aimed primarily at investigators of psi claims and admirably serves their purposes. On the other hand, it might well be valued, also, by those wishing to successfully produce fraudulent effects. The chapter headings read like a set of condensed instructions on how to succeed at faking psi events. This might, ironically, provide more targets for exposes. The following are the first of the superordinate chapter headings and its subordinate headings: Misframing, followed by, Appear Incapable of Fraud, Appear to Have No Motivation for Fraud, Appear to Be Unwilling to Engage in Fraud, Create a Believable Claim, and Produce a Claim that the Individual "Wants" to Believe, (pp. 36-41). To further capture the flavor, savor the remaining set of superordinate headi ngs: Hinder the Development of "Normal" Explanations, Manipulate an Individual into Incorrectly Believing Normal Explanations Implausible, Exploit Ineffective Controls, Exploit the Conditions Apparently Needed to Elicit Psi, and Have "Outs" Ready in Case Something Goes Wrong (pp. 42-50). This material should be read by anyone wishing to investigate (or evaluate reported investigations of) those claiming to be psychics. Readers should not, though, expect to be told how specific pseudopsychic tricks are done. This chapter is concerned only with operational principles that could help to ensure that the deployment of deceit truly is effective.

The authors state that this chapter "outlines the beginnings of a cognitive model which identifies the main stratagems of psychic fraud" (p. 36). The information in this chapter seems, though, more like raw material for this purpose, because its level of discourse hardly seems abstract enough and because the structural and functional features associated with traditional cognitive models are essentially absent. There is, for example, not a single diagram. Nonetheless, this essay is a thoughtful and highly useful one for the audiences mentioned earlier.

Chapter 3, "The Feilding Report: A Reconsideration," may be the most likely of the chapters (material first appeared in the Journal of the Society for Psychical Research, 1992) to arouse the ire of parapsychologists who cherish ostensible psi events of high magnitude and who have regarded the Feilding Report (Feilding, Baggally, & Carrington, 1909) as providing some of the best evidence for the alleged phenomena of physical mediumship. Wiseman has found evidence of important inconsistencies in the report and has pinpointed ambiguities and vagueness. His writing, although usually restrained and fair, at times lapses here into a tone that might be seen as accusative or prosecutorial.

Wiseman assuredly has contributed some valuable insights and commentary. One of the most important is his pointing out the bothersome disparity between various published diagrams of the seance room and what is shown in three photographs of that room that are contained in the original report. The photographs show the seance curtain stretched across a back corner of the room and positioned so that it obscures vision of much of the right side of a double door to an adjoining room. The published diagrams, on the other hand, unlike one in the Society for Psychical Research (SPR) archives, show the curtain hung so that no obscuration of this door would be possible. It is through a rigged lower panel on the obscured portion of the double door that, Wiseman suggests, an accomplice might have entered and left the room during seances.

This disparity of photographs and published diagrams is troubling for reasons that go beyond any ambiguity the disparity may introduce about the paranormality of events in the seance room. Why should the published diagrams be different from what is shown in the photographs and different from a sketch in SPR archives, different in a way that could make the chances of medium fraud appear less than they might have been? We do not know the answer to this question, and, so far as I can ascertain, it can only be answered speculatively. Speculation, of course, often reflects the speculator's biases. An answer to this question is not necessary, though, to recognize that Wiseman has pinpointed a potential threat to a paranormal interpretation of the findings in the Feilding Report. Although some investigators were seated so that they might have glimpsed movement of an accomplice crawling between the lower door panel and the seance curtain, their seating arrangement meant that their backs were turned on this potential trouble spot.

Wiseman has read the paper like a trial attorney trying to tear apart a case, and that can be useful in an instance such as this. That having been said, Wiseman's critique is in a scientific journal, not in a courtroom proceeding where the language of rhetoric sometimes exceeds that of reason. He has, from time to time in this chapter, lapsed into a style that exceeds the evidence, that seems unfair, and that does not fit well with the objectivity required of the scientist. One example: "Yet the investigators choose not to inform their readers whether or not such a trap door was possible in the Hotel Victoria, Naples" (p. 82, my emphasis). My quarrel here is with the allegation of what could seem, given the wording, like a deliberate act of deceit by the reporter-investigators, a kind of gilding of the lily on their part. It would have been far more polite, less accusative, and less supposititious to say, "Yet the investigators did not discuss whether or not such a trap door was possible in the Hotel Victori a, Naples." Perhaps Wiseman wished to be accusative. His judgment of the intellectual integrity of the Feilding Report authors might have been seriously compromised by the important diagram-photo disparities discussed earlier. Nonetheless, whatever his personal opinions, his work might have been more effective among parapsychologists-who should be its chief important audience, unless he wishes to preach to the converted (i.e., to other skeptics)--if it had adopted a less accusatory tone. Scientific critiques should concern facts, not accusations concerning motivations. Although it may be necessary to discuss motivations in certain instances, this should proceed with the greatest care, caution, and fairness.

Aside from the norms of fairness and collegiality, these occasional lapses in Wiseman's writing here are a cause of concern because they almost always are unnecessary in order to point out weaknesses in the investigation and/or in the report. His lapses in this regard might offend some parapsychologists to the degree that they would tune out his very important messages. That would be most unfortunate. It is easy to slip over from exposing circumstances that need to be exposed to making explicit or implicit accusations about the motives, intentions, or intelligence of investigators. A clear case of fraud or deceit is fair game, but all critics, including self-professed skeptics, should remember that the target of scientific writing ordinarily should be the circumstances that bear on the correctness of investigators' conclusions, not the motivations or intelligence of those investigators. Wiseman's record as a skeptic is a good one, as discussion in this review should make amply clear, and one hopes that he wi ll maintain that good record by keeping his discourse free of unnecessary hurt to those whose work he is criticizing. As work reported in his book amply shows, Wiseman is acutely aware of how the beliefs or schemata of individuals can bias their cognition. Therefore, he should know that expectation-driven lapses of judgment, perception, and memory occur and can produce deficiencies in scientific investigation that cynics might wish to explain by duplicity or intellectual deficit.

In terms of substance, Wiseman's Feilding Report critique seems better balanced than much of the past criticism of other reports by other critics, and it is somewhat less accusative in tone. As an example of the former, Wiseman applies his wonderfully analytic mind to question specific prior efforts to criticize the conclusions of the Feilding Report.

Because it is not always easy to discern problems under the pressure and exigencies of the planning or execution phases of research, all research warrants careful retrospection. Wiseman's critical work on the Feuding Report is, in my view, a worthwhile example of an outsider doing precisely that, excepting his occasional lapses into an accusative mode.

Chapter 4, "Testing the ESP Claims of SORRAT" by Wiseman, John Beloff, and Robert L. Morris (material first published in the Journal of the Society for Psychical Research, 1992) may elicit a yawn or two from readers, as it did for me. It simply is not very interesting, for some of us, at least, to read about farfetched claims and to learn that they have not been supported by careful research. On the other hand, a psi skeptic might have particular interest in sensational claims of alleged psi events, such as those of The Society for Research in Rapport and Telekinesis (SORRAT), a spiritualist group in Missouri. A large variety of startling claims, of both the psychokinetic and extrasensory kinds, have been made by SORRAT on behalf of supposed spirit entities.

These authors' test of SORRAT claims about extrasensory card guessing is perhaps most intriguing as an essay in incredibly elaborate and burdensome precautionary methodology. It failed to produce a statistically significant outcome. Another negative outcome was a failure to have psychically rearranged the cards of the test deck, which the "entities" had claimed they had done.

A unique element in this report is that Ed Cox, who was an on-site facilitator in this work, and Tom Richards, one of the latter-day leaders of SORRAT, were asked "to approve of and sign an agreement that clearly outlined the experimental protocol" (p. 115). The purpose of this was stated as an effort to obviate any subsequent claims that "the experimental conditions were not favorable to the production of psi" (p. 115). Among other things, this agreement clearly stated what would happen if the experiment were successful in terms of correct guessing, in the absence of signs of the protective package having been damaged or otherwise violated. The authors would then state that "we are impressed by the results, that we have no ready conventional explanation for the results (sic, omitted comma) and that we feel further research is warranted" (p. 122). The signed agreement included, "However, we will not issue any statement to the effect that we believe the claimant to have 'psychic' ability" (p. 122). It is good that the participants should know this in clear and unequivocal terms.

Fair enough, but this agreement still seems a bit incomplete. Should not something be said in such agreements about what would be published should there be a failure to find evidence of paranormality? I am not suggesting any constraint on what the investigators should say about negative results, if they should occur. My suggestion is that investigators should make this very clear to participants before beginning research with them. A psychic claimant, even a fully honest one, might want to demand such a statement because scientists sometimes overgeneralize or overstate the implications of their results. The actual agreement made minimal concessions if the study should be successful, and it apparently did not ensure the research participants against over-general conclusions if it should not be. (This is not at all to suggest that these investigators would have overstepped legitimate boundaries, only that, as a general practice, it would be good to clarify these matters for research participants before they be gin participation.)

The SORRAT chapter has substantial value in providing guidance about needed precautions in research of this kind. This study, nonetheless, took much for granted, psychometrically speaking, because it imposed a very high performance level (60% where 20% is chance) as a prerequisite for judging the study a success, and the reliability of measurement was dubious because of a single test run having been successfully done. Given the difficulty of getting ESP-target calls from SORRAT, as evidenced in this study, given the time and effort required to impose conditions like these, and given the poor performance, the lack of additional testing is understandable.

"An Experimental Test of Psychic Detection," Chapter 5 (material first published in the Journal of the Society for Psychical Research, 1996), finds Wiseman joined by Donald West and Roy Stemman. This study tested the abilities of "three well-known psychic detectives" (p. 147), as well as of a group of university students who had made no claims about being psychic and who professed no particular interest in criminology. The purpose was to see, using a paradigm similar to that often used by "psychic detectives" (e.g., handling crime-related objects), whether either group was able to discriminate statements related to a given crime from those that were not related to it, with each statement being related to one of three actual crimes. This study is both interesting and troublesome. It is interesting because it addresses an important issue in a generally thoughtful way. It is troublesome because the methodology might not have supported a firm conclusion about psychic detection, even had there been statistical si gnificance, which there was not. The major positive conclusion the authors reach as a result of this study is that psychics make more statements in ruminating on a crime than do students. This and other considerations suggest to me that signal detection methodology might profitably be used in studies of this kind.

A final comment worthy of consideration is that the crimes in this work already had been solved. Many psychics claim and apparently believe that real needs are what drive their ability, but a vital need is not present in the case of a study with solved crimes. One might hope, next time, for a proactive study, even if the results might not be known for some time. The study would, of necessity, have to use a different methodology than that outlined here.

"Investigating Macro-PK in India: Swami Premananda," Chapter 6 (material first published in the Journal of the Society for Psychical Research, 1995), co-authored by Erlendur Haraldsson, reports an investigation of alleged PK by a well known Hindu religious leader. Among this swami's claimed paranormal manifestations are the regurgitation of Shiva lingame (i.e., small egg-shaped stones) and the materialization inside his bands of vibuti (i.e., sacred ash) and various small objects. It is good to find Wiseman paired with Erlendur Haraldsson, a parapsychologist who certainly is not hostile to the possibility of paranormal events around religious figures and who had previously studied Premananda in a less formal way with inconclusive results. It is my impression from reading the chapter that Haraldsson contributed very substantially to the psychological atmosphere that helped to sustain Premananda's cooperativeness, even after initial failure. This chapter has, in my view, considerable interest despite the fact that Premananda did not succeed in producing materialization under stringently controlled and adequately videotaped circumstances. The nuances of social interaction between this religious figure and these scientific investigators make fascinating reading. So, also, do the efforts of this swami to be cooperative, to try to make things happen under stringent control. Vibuti was "materialized" by Swami Premananda, albeit not under conditions where sound conclusions could be drawn.

It is exciting to see a skeptical investigator with training in legerdemain paired with an optimistic parapsychologist thoughtfully attuned to the psychosocial needs of research participants who are accustomed to being treated with deep respect and even veneration. We need more individuals trained in magic who are willing to investigate in such ways and more parapsychologists who are willing to work with them. Such developments are salutary and bode well for future work with those who claim strong psi manifestations.

Haraldsson and Wiseman paired up a second time for "Assessing Filmed Evidence of Alleged Trickery by Sathya Sai Baba," Chapter 6 (material first published in the Journal of the Society for Psychical Research, 1995). Haraldsson had earlier published a book on the psychic claims of Sai Baba, a very well-known Hindu religious figure in India. Sai Baba's claims to fame include numerous and frequent alleged macro-psi events. The present chapter represents these authors' efforts to examine for evidence of fraud a video sequence of Sai Baba, alleged by the Deccan Chronicle, an English-language newspaper in Hyderabad, to show fraud. Despite these investigators' assiduous efforts to enhance the image quality of the video sequence, no conclusion of fraud was possible on the basis of the tape. A very refreshing statement by Haraldsson and Wiseman is dished out in the very last sentence: "This is a clear example of how the press can alter ambiguous evidence of possible fraud into definite proof of cheating" (p. 195). Di d you think you would ever hear words like that from a self-professed skeptic? These are among other signs, some of them evident in this book, that Wiseman's orientation is unlike that of the typical skeptic.

Wiseman joined Haraldsson once again in "Three Further Investigations of Ostensible Macro-PK in India," Chapter 8 (material originally published in Journal of the Society for Psychical Research, 1996). This, like the earlier work with Swami Premananda, provides fascinating reading, although no satisfactory evidence of macro-PK emerged in any of the three investigations. In two of the three studies, these investigators found what they deemed to be evidence of trickery, and in one study, it looked to them like crude trickery, not even minimal sleight of hand. In none of these three studies could the participants be called highly cooperative. These investigators' luck was a bit better, though, in the second investigation, in that the claimant there somehow produced small amounts of vibuti-like power atop two picture frames and did so under conditions not deemed by the investigators to be particularly suspicious. (The unanticipated episode was not videotaped because a demonstration had not been expected, and the investigators were caught by surprise.) Haraldsson and Wiseman had inspected the frames shortly before the alleged materialization, but after this inspection the claimant had handled the pictures prior to the "materialized" substance being found. The claimant was seen to place the pictures directly under a large fan before going into meditation in an apparent effort to produce a materialization. Haraldsson and Wiseman, by their own statement, made two serious errors in this case: They failed to inspect the picture frames after the claimant had handled them but prior to the claimed materialization, and they neither collected the "materialized" substance nor carried with them collecting tools or a receptacle suitable for such material. Because this claimant twice turned down requests for formal demonstrations but did one under his own conditions, I find this episode less interesting than do, apparently, the authors of this report. I am a bit concerned by what seems like a lapse of criticality in this investigation and report. How so?

The morning after this demonstration the claimant visited the investigators in their hotel room, but he did not accept their invitation to do another demonstration. (He also had refused their invitation to demonstrate earlier on the day of the actual demonstration, the latter not having been at their request.) He did give Wiseman, on the morning following the demonstration, what was said to have been one of the framed pictures used in it. The investigators stated that the claimant "gave Wiseman one of the pictures used during his demonstration and departed" (p. 207). Considerable is made of this in the report because three subsequent chemical tests were done on the picture's frame, none with clearly affirmative results, to test nonpsi chemical hypotheses about possible origins of the powder observed atop the photographs at the time of the demonstration.

It seems reasonable to ask how these investigators could be sure that the frame given them by the claimant actually was the same one used in the demonstration or, barring substitution, whether its surface might somehow have been altered in the interim. The substitution possibility may be the more plausible of the two. Neither possibility seems more farfetched than some that Wiseman entertains in criticizing the Feilding Report nor than things he suggests psi claimants might do in other situations. Haraldsson and Wiseman, nonetheless, appear to congratulate themselves about their accomplishment relative to getting the claimant to give them (what was allegedly) one of the two pictures used in the demonstration of the preceding day: "The fact that Wiseman managed to acquire one of the pictures used during the demonstration allowed us to retrospectively assess some possible 'normal' (i.e., non-psi) explanations for the phenomena (sic, so far as I know only one event or class of event was involved)" (p. 213) and "This illustrates the potential importance of investigators (sic, omission of an apostrophe) obtaining the important objects involved in a demonstration" (p. 213). Getting such objects (including the allegedly materialized material or object) is very important, but in this case the investigators did not get the latter and conceivably might not have got the former. Wiseman was very meticulous in examining possibilities of claimant deceit in a number of his other papers in this book, so his lapse in this report is surprising. Also, the authors did mention that the claimant had elected to place the pictures, during the demonstration, underneath a large fan, but they did not discuss the possibility that the "materialized" material might simply have been fan debris. I occasionally find gray fan debris, which sometimes 'sticks up like brisdes" (p. 207), underneath fans that have been running but that have not been cleaned for some time (and this was mid-July in India). Their remarks, on the other hand, might be see n indirectly to suggest that the fan conceivably could have been used to help deposit supplied material onto the frame.

The bottom line here is that although these authors are admirably frank in admitting mistakes, those who wish to critically evaluate claims of materialization should come expecting impromptu demonstrations and be as ready as possible to carefully observe and collect relevant physical evidence. (It would have been very interesting had the vibuti-like material turned out to be fan debris.) Relying on memory, as they did, about the nature (e.g., color and structural appearance) of the materialized substance is highly unreliable. We are not told whether on-site notes were made. Even having used a sheet of paper to brush the vibuti-like material onto another sheet, the latter folded for safekeeping, would have been far better than nothing. Investigators must also ensure that the evidence they collect is the true evidence, not surrogate evidence. Wiseman well understands the latter point, as is clear from examples elsewhere in this volume. Impromptu demonstrations were the rule, not the exception, in their work on alleged macro-PK in India. "Be prepared" is the Boy Scout oath, and it also should become that of field investigators of alleged macro-psi. Another thought: Are skeptics sometimes so focused on human deceit (e.g., sleight of hand) that they fail to gather evidence, oradequate evidence, to allow their study to examine other hypotheses (e.g., the possibility of genuinely inexplicable events, as in the chapter on extrasensory detectives, or normal events, as in the case of fan debris)?

For the psychologically oriented reader, the last three chapters (9, 10, and 11) of this volume maybe the most interesting. The first of these ("Recalling Pseudopsychic Demonstrations," coauthored by Wiseman and Robert L. Morris; material first published in the British Journal of Psychology, 1995) examines how the sheep/goat (psi believer/psi nonbeliever) distinction relates to subsequent memory for a pseudopsychic demonstration. The second ("The Effect of Belief in the Paranormal and Prior Set upon the Observation of a 'Psychic' Demonstration," by Matthew D. Smith and Richard Wiseman; material first published in the European Journal of Parapsychology, 1992, but modified here) examines, jointly, psi belief and observational set (i.e., is the feat stated to be "authentic" or "trick") in relation to recognition memory for events that happened during such a demonstration. The third ("Manifestations: An Experimental Investigation into Seance Room Phenomena," by Richard Wiseman, Jeff Wiseman, and Matthew Smith; m aterial not previously published) examines memory for events during a seance-like setting. Readers of these chapters may find, as I did, that the results of these studies are disappointing in their lack of definitive, conceptually clear outcomes and in terms of how, in a number of instances, the studies were conducted and analyzed. These three chapters are, despite these problems, well worth reading because they involve innovative efforts to address problems of interest both to parapsychology and to cognitive psychology. Most of my remarks here will focus on the first two of these chapters.

The Wiseman-Morris report (Chapter 9) disappoints partly because it is incorrectly stated that the study examined "recall." In fact, the retrieval measure was a recognition task, one in which participants read four statements about the content of the videotape and rated the degree to which each was felt to be true or false. Some were true and some were false. This clearly involves a recognition task with foils, combined with a confidence rating. It resembles the kind of recognition task that sometimes has been used in studies of recognition memory that involve signal detection methodology, although its results were scored very differently.

Very important, the methodology confounds true recognition performance with response bias (or criterion-related effects). Response bias, which is strongly influenced by expectations, can determine the magnitude of the rating. Given that examining the role of expectations in relation to encoding was a major objective of this work, it is very unfortunate that the study confounds these two matters. Signal detection methodology, which was not used here, separately considers both discriminability (i.e., the ability to discriminate old from new stimuli in the case of recognition memory work) and response bias or criterion (i.e., roughly, the readiness to report a given class of stimuli). The very great importance of differentiating these factors in work on expectancy-congruence/incongruence in memory has been amply shown in meta-analytic work by Stangor and McMillan (1992) and, in the case of schemata and memory, meta-analytic work by Rojahn and Pettigrew (1992). Both of these very important papers should be consu lted by anyone planning work in this area. Work that confounds discriminability and response bias, as does the Wiseman-Morris work, is likely to produce minimally interpretable or misleading outcomes.

The remarks that follow in this section (and beyond this paragraph) refer to outcomes during the "First Recall Period" for both studies. For brevity, I have not discussed results for the "Second Recall Period," although similar considerations are relevant there. Another concern about this work is that much is made of significant differences between sheep and goats on the important items of the post-video memory measure. ("Important items" were those independently rated by two magicians as being important to the methodology of a trick. "Unimportant items" were those not thus rated by the magicians.) Goats tended to do better on important items than did sheep. Unfortunately for the authors' preferred interpretation of these data, the interaction of the sheep-goat and the importance variables apparently was not tested (nor were related, hypothesis-implied, cross-importance contrasts at a given level of belief). Doing these analyses would have addressed whether participants' beliefs made the important items spec ial, an apparent implication of the investigators' hypothesis. Perhaps most important, it was not shown that the goats scored higher on the important than on the nonimportant items. Given the information provided in the tables and that the importance variable was within-subjects, I was unable to compute inferential statistics for the relevant contrasts. Analyses were, in short, not done (or were not reported) that were critical to testing implications of ideas underlying this work. These remarks apply to both pseudo-PK and pseudo-ESP demonstration results.

In planning Experiment 2, the investigators did consider the response bias problem in that they included pseudo-important questions in the post-video questionnaire and refocused their interest largely on them, as contrasted with the important questions. This was a step forward, although I doubt that it adequately addresses the response bias issue. At any rate, there was some inconsistency in positive findings across the two studies, and, embarrassingly for their hypothesis, on the unimportant questions, goats scored better than sheep, either suggestively (.06, two-tailed, pseudo-PK demonstration) or significantly (.005, two-tailed, pseudo-ESP experiment). I really do not know what to conclude from this work except that its tantalizing, but very mixed, outcomes with inadequate memory measures warrant further investigation with more adequate methodology.

The authors are to be commended on their own critical observations about their own work, as expressed in their General Discussion section. They did not, however, cover the problems discussed in this review.

That sheep did not significantly rate demonstrations (prior to debunking) as more paranormal than did goats, in 3 of 4 sub-studies, might have been due to impression management. Students might have been aware of the possibility of skepticism by the experimenters and might have wanted to avoid looking foolish. These results should perhaps, then, be taken with a grain of salt. This possibility of an inadequate test of their prediction merited discussion.

Finally, in work of the Wiseman-Morris kind, it would be good to examine whether goats might show their own kind of cognitive error. Might they believe that they remember things indicating a nonpsi explanation when those features were not present in the actual demonstration? (This might be particularly evident with a long retention interval.) Although Wiseman and Morris discuss the importance of future work to examine possible biases by goats, their remarks focus on attribution when observing a demonstration, not on possible false memories on the part of goats. This false-memory issue is important because a number of skeptics have attempted to debunk performances by alleged psychics by saying, afterward, that they saw specific, concrete indications of trickery (or, at least, saw efforts to use trickery).

By contrast, the Smith-Wiseman study in Chapter 10 may prove of relatively little interest to readers, not because it addresses an uninteresting topic (recollection as influenced by information about the authenticity of a psychic performance and the role of belief in the paranormal in possible interaction with the preceding manipulation), but because the only effect was a main effect of the experimental manipulation and because the dependent measure (Psychic Surgery Questionnaire, PSQ) seems unlikely to effectively test these ideas. As the authors themselves noted, the six items of the questionnaire "were not particularly relevant for the accomplishment of trickery" (pp. 248-249). I would note that the details addressed by these questions seem as if they might have been pretty much irrelevant to anything of real interest in the film, at least for typical observers, and they seemed to have involved information that was simply unlikely, in many instances, to have been processed effectively.

There appears to be a minor error in the analysis of variance summary table (Table 10.3) for this study. The F-ratio associated with the sheep/goat variable should have been 1.57, not 1.33 as reported, although this should not have affected statistical inference. In their "Commentary on Design and Methodology" section, Smith and Wiseman discuss some much more important problems that they see in their study. One of them concerns that of dichotomizing participants as sheep and goats and the fact that median splits can produce meaningfully different splits across studies. They suggest, instead, a trichotomization of participants with the middle third being dropped from the analysis. This does not seem a satisfactory resolution, and it has a similar problem across samples as does dichotomization. What Smith and Wiseman do not focus on is that underlying this difficulty is the central problem of loss of information through dichotomization (or trichotomization) of continuous variables. A more adequate solution tha n theirs may be the use of multiple regression techniques, which allow continuous variables to be treated continuously. It would be advisable, before using that approach, to ensure that the regression of the dependent measure on the belief measure is essentially linear in form, given that the technique operates off this assumption. If linearity of regression can be established, this technique should be more powerful than with the loss of information typically found in such work. If a scatterplot should show a simple, nonlinear regression, namely a v-shaped or u-shaped function, this could be very important in understanding what is happening. For example, such curves may combine two genuinely separate functions.

A major problem in the Smith-Wiseman work is one that I do not recall as having been acknowledged by them. It concerns the difficulty of drawing conclusions on the basis of performance of sheep and goats on some dependent measure. Is the difference due to belief in psi or due to skepticism concerning it? Comparing only sheep and goats can provide no answer. This is another possible reason for not dropping out the middle third of the sheep-goat distribution. If one wishes to trichotomize, the middle group can serve, potentially, as neutral comparison group, but only if the distribution makes it truly neutral. (Multiple regression can also be used to examine these issues, but this is not the place to discuss that.)

Ironically, an example of possible expectation-related memory bias on the part of these authors occurs when they, in referring to the Wiseman-Morris work just reviewed, say "Wiseman ... found that those subjects designated as sheep rated the performances as more paranormal in content than did the goats" (p. 244). What Wiseman and Morris found in three out of four cases was nothing even approaching significant evidence for this effect, despite a one-tailed test. Nor did they report a meta-analysis across subsamples to show that significance held across them.

A final concern about the Smith-Wiseman paper, which actually was somewhat modified from the earlier l992 journal paper, is the wording of the title. It includes "The effect of belief...." These reporters would seem to share an error extremely common in contemporary psychological reports, including in convention papers and posters and in published materials. This is that of using causal language to describe correlational findings. How, indeed, does one infer causation on the basis of a set of prior beliefs about psi reality, which certainly are not experimentally manipulated? This is the more difficult because, as Wiseman and Morris noted in the chapter just discussed, prior work has shown that belief in the paranormal correlates with measures of cognitive ability. Consequently, differences in cognitive ability and/or style may affect beliefs about paranormality, and the same cognitive differences may affect performance on the dependent measures used in studies such as these. Belief may not itself be the cau sal factor. Conceivably, it may not even be a mediator variable, or, of course, it may be one or both. Causation-related issues in correlational domains such as these must be examined through very careful and thoughtful theory-driven analysis with modern analytic tools such as path analysis and structural equation modeling (see, e.g., Kenny, Kashy, & Bolger, 1998).

Although the final report, by Richard Wiseman, Jeff Wiseman, and Matthew Smith, is intriguing because it examines memory for discrete events occurring in a seance-like setting, it has not been published and includes no statistical evaluations, possibly due to a possible lack of independence of observations in this group seance setting. Because it follows two fairly heavy-duty experimental-statistical papers on cognition related to psi beliefs, this nonexperimental, nonstatistical study is a bit like an intriguing and tantalizing, but in the end not terribly fulfilling, dessert coming at the end of a hefty meal. It does, though, provoke thoughts about follow-up work.

This book lacks an index, which makes it less useful than it might have been. The lack of an editor also proves a problem because Wiseman's writing, although generally very clear, sometimes becomes overly repetitious or needs an occasional correction of punctuation. Real deficits and annoyances in the book are the use of old-fashioned numbers in lieu of surname-date citations (e.g., Wiseman, 1999) and reference listings accordingly, namely a separate reference (albeit partial for successive citations) for each such citation. This is bothersome to the reader, who sometimes cannot identify a source by looking at the numbered reference and has to waste time scanning up and down a page and within hard-to-read multi-reference notes! This also wastes paper (hence, trees) and raises book costs.

In the end, the reader may be disappointed by the lack of clear sets of positive findings that emerge from this volume, and he or she may even be bothered by overstatements about the outcomes of particular studies or by somewhat exaggerated claims regarding the value of methods introduced, here and there, by Wiseman. On balance, though, that is not my predominant reaction. Wiseman's writings in this volume and elsewhere would seem to represent a somewhat different variety of criticism than has historically been found in the writings of self-professed skeptical authors. These reports seem, for the most part, to be more constructive and fair in style and in spirit than what we often have seen from skeptical individuals. A very large plus is that Wiseman himself is a very active investigator.

Parapsychologists have sometimes complained that skeptical authors should get out of their armchairs and do some first-hand investigation. Wiseman is no armchair critic, and he actively investigates both psi-related possibilities and the possibilities of deception and self-deception that sometimes muddy the waters in this field. His comments on his own work are refreshingly forthright and thoughtful in pointing out major needs for improvement. This is true even though his methodology and citations, when he investigates psi beliefs and cognition in this volume, seem less than adequately informed by the conceptual and empirical literature on expectancy and memory already in existence when he presumably did this work and prepared it for publication.

Wiseman's work and writings, as evidenced in the present volume and elsewhere, should, in my view, aid parapsychology in (a) effectively investigating psi claims, especially those related to macro-psi, and in (b) advancing the understanding of psychosocial factors that contribute to erroneous perceptions and recollections of demonstrations of ostensible psi. It is hoped that the suggestions and criticism in the latter part of the present review might help to advance the second objective. The latter domain, if developed properly, should make useful contributions to cognitive and to social psychology. Science is advanced by those who directly investigate and who contribute thoughtfully and fairly, by commentary and by investigation, to the evaluation of earlier claims in a given field. Many parapsychologists likely will disagree with some of Wiseman's conclusions about particular works of others, and some with his conclusions from his own work, but I doubt that many, if any, would deny that he already has cont ributed valuably to this field through both his criticism and his investigations. If you have not already read most of the studies in this volume, you might profit by so doing. What is most clear is that we now have a vigorous and thoughtful skeptic actively contributing to this field, one who prefers investigation to pontification.

(1.) I thank Vandana Rajpal for reading the first draft of this paper and for making many valuable suggestions.


FEILDING, E., BAGGALY, W. W., & CARRINGTON, H. (1909). Report on a series of sittings with Eusapia Palladino. Proceedings of the Society for Psychical Research, 23, 309-356.

KENNY, D. A., KASHY, D. A., & BOLGER, N. (1998). Data analysis in social psychology. In D. T. Gilbert, S. T. Fiske, & C. Lindzey (Eds.), The handbook of social psychology: Vol. I (4th ed., pp. 233-265). New York and Oxford: Oxford University Press.

ROJAHN, K., & PETTIGREW, T. F. (1992). Memory for schema-relevant information: A meta-analytic resolution. British Journal of Social Psychology, 31, 81-109.

STANGOR, C., & McMILLAN, D. (1992). Memory for expectancy-congruent and expectancy-incongruent information: A review of the social and social developmental literature. Psychological Bulletin, 111, 42-61.
COPYRIGHT 2000 Parapsychology Press
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2000 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Review
Publication:The Journal of Parapsychology
Article Type:Book Review
Date:Jun 1, 2000

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters