ArticlePDF Available

Magic and memory: Using conjuring to explore the effects of suggestion, social influence, and paranormal belief on eyewitness testimony for an ostensibly paranormal event

Authors:

Abstract and Figures

This study uses conjuring to investigate the effects of suggestion, social influence, and paranormal belief upon the accuracy of eyewitness testimony for an ostensibly paranormal event. Participants watched a video of an alleged psychic seemingly bending a metal key by the power of psychokinesis. Half the participants heard the fake psychic suggest that the key continued to bend after it had been put down on a table and half did not. Additionally, participants were exposed to either a negative social influence (a stooge co-witness reporting that the key did not continue to bend), no social influence, or a positive social influence (a stooge co-witness reporting that the key did continue to bend). Participants who were exposed to the verbal suggestion were significantly more likely to report that the key continued to bend. Additionally, more participants reported that the key continued to bend in the positive social influence condition compared to the other two social influence conditions. Finally, believers in the paranormal were more likely to report that the key continued to bend than non-believers.
Content may be subject to copyright.
ORIGINAL RESEARCH ARTICLE
published: 13 November 2014
doi: 10.3389/fpsyg.2014.01289
Magic and memory: using conjuring to explore the effects
of suggestion, social influence, and paranormal belief on
eyewitness testimony for an ostensibly paranormal event
Krissy Wilson and Christopher C. French*
Anomalistic Psychology Research Unit, Department of Psychology, Goldsmiths, University of London, London, UK
Edited by:
Jay Olson, McGill University,
Canada
Reviewed by:
Richard Kanaan, Institute of
Psychiatry, UK
Fiona Gabbert, Goldsmiths,
University of London, UK
*Correspondence:
Christopher C. French, Anomalistic
Psychology Research Unit,
Department of Psychology,
Goldsmiths, University of London,
Lewisham Way, New Cross,
London SE14 6NW, UK
e-mail: c.french@gold.ac.uk
This study uses conjuring to investigate the effects of suggestion, social influence, and
paranormal belief upon the accuracy of eyewitness testimony for an ostensibly paranormal
event. Participants watched a video of an alleged psychic seemingly bending a metal key by
the power of psychokinesis. Half the participants heard the fake psychic suggest that the
key continued to bend after it had been put down on a table and half did not. Additionally,
participants were exposed to either a negative social influence (a stooge co-witness
reporting that the key did not continue to bend), no social influence, or a positive social
influence (a stooge co-witness reporting that the key did continue to bend). Participants
who were exposed to the verbal suggestion were significantly more likely to report that the
key continued to bend. Additionally, more participants reported that the key continued to
bend in the positive social influence condition compared to the other two social influence
conditions. Finally, believers in the paranormal were more likely to report that the key
continued to bend than non-believers.
Keywords: magic, memory, suggestion, social influence, paranormal belief
INTRODUCTION
For centuries, magicians have amazed audiences by apparently
defying the laws of nature. Such effects were based upon a
deep understanding of lay psychology but until recently, with
few exceptions, academic psychologists have largely ignored the
insights that the art of conjuring can provide to help understand
the workings of the human mind. Thankfully, as this special issue
demonstrates, this situation is changing. One of the ways in which
the art of conjuring can be of service to psychological science is
by providing means to study a range of psychological phenomena
such as perception and memory. The experiment described in this
report is one such example.
Over several decades, a great deal of research has demonstrated
the unreliability of memory and in particular the fallibility of
eyewitness testimony. Many kinds of memory distortion effects
have been investigated including those due to the presentation of
post-event misinformation (e.g., Eakin et al., 2003) and the use of
misleading questions (e.g., Loftus, 1975) and even the formation
of detailed false memories for complete episodes (e.g., Loftus and
Pickrell, 1995). Recently, researchers have turned their attention
to a particular form of misinformation effect known as memory
conformity (e.g., Wright et al., 2000, 2009; Gabbert et al., 2003;
Gabbert and Hope, 2013). Memory conformity is said to occur
when an individual memory report of one person becomes more
similar to another person’s following their discussion of an event.
In forensic contexts, similar accounts from multiple wit-
nesses are likely to be accorded greater evidential weight than
an uncorroborated account from a single witness. While such
an assumption may be defensible, it fails to recognize that mul-
tiple witnesses to an unusual event such as a criminal act are
very likely to discuss the event before any formal investiga-
tion takes place. Information exchanged during such discussions
may potentially change or add to original recollections of what
happened. Researchers have investigated how memory recall of
pairs of eyewitnesses can become distorted if the two witnesses
discuss what they believe to be the same event. Gabbert et al.
(2003) had pairs of participants watch a video of a staged crime
recorded in such a way that crucial details that were available
on one recording were not available on the other and vice
versa. For example, one version of the video showed a young
woman actually steal some money whereas it was not clear in
the other version if she had done so as it was filmed from a
slightly different viewpoint. Dyads in one condition discussed
the event prior to recall while participants in a control condition
did not. It was found that a significant number of participants
erroneously included items of information in their report of
the event that had been acquired as a result of discussion with
a co-witness. For example, many of the participants who had
not actually seen the young woman take the money mistak-
enly reported that they had seen this act following the dis-
cussion. These findings were replicated by Wilson and French
(2004).
In addition to reports of criminal acts, the accuracy of eyewit-
ness testimony is also crucially important in assessing reports of
ostensibly paranormal experiences (OPEs) and other anomalous
events (French, 2003; French and Wilson, 2006). Anomalistic
www.frontiersin.org November 2014 | Volume 5 | Article 1289 |1
Wilson and French Magic and memory
psychologists have argued that most reports of OPEs can be
plausibly explained in non-paranormal, typically psychological,
terms and specifically that cognitive biases known to characterize
human thought may lead many people to believe they have
experienced something paranormal when in fact they have not.
Although a wide range of cognitive biases are potentially of
relevance in this regard (French, 1992; French and Wilson, 2007;
French and Stone, 2014), memory-related biases are amongst the
most important. French (2003) and French and Wilson (2006)
presented comprehensive reviews of investigations of the accuracy
of eyewitness accounts of OPEs, concluding that anecdotal reports
of such events should be treated with considerable caution in light
of the proven unreliability of memory in such circumstances.
Wiseman and Morris (1995), for example, compared the recall
of believers and disbelievers in the paranormal for the details of
pre-recorded “pseudo-psychic” demonstrations, such as apparent
metal-bending by psychokinesis. Believers tended to have poorer
recall of the details of the demonstrations, particularly those
details that would give some indication of the type of sleight of
hand that was used to achieve the effects. Perhaps not surprisingly,
the believers rated the demonstrations as being more “paranor-
mal” than disbelievers.
Poor recall of the events taking place in séances was demon-
strated as long ago as 1887 by Hodgson and Davey (1887),
with similar findings being reported by Besterman (1932) and
more recently by Wiseman et al. (1995). In all such studies, all
of the effects were achieved by the use of trickery based upon
accounts from fake mediums. However, the accounts provided by
eyewitnesses were often so inaccurate that, taken at face value,
they would defy rational explanation. Once again, important
details of the events that would have provided clues as to how
the effects actually had been achieved were simply not recalled
accurately.
Wiseman et al. (2003) examined the effects of suggestion
during fake séances. In their first experiment, around a third of
the witnesses erroneously reported that a stationary table had
moved during the séance following a suggestion from the fake
medium to this effect. Believers in the paranormal were more
likely to misreport such movement than disbelievers. Believers
were shown to be more susceptible to suggestion than disbelievers
in a second set of fake séances too, but only when the suggestion
was congruent with their belief in the paranormal. For example,
if the fake medium suggested that an object had not moved when
in fact it had (by trickery), believers were no more likely to accept
the suggestion than disbelievers. Overall, around one-fifth of the
participants believed they had witnessed genuine paranormal
phenomena. As Wiseman et al. (2003) point out, it is unclear
whether the verbal suggestion directly affected the participants’
perception of the event, their memory of the event, or both. It is
even possible that neither perception nor memory was affected
and that the results were due to demand characteristics, but the
end result is the same: a large minority of the participants were
willing to report that stationary objects had moved and that they
had witnessed genuinely paranormal events.
Wiseman and Greening (2005) explored the power of verbal
suggestion in another ostensibly paranormal context. In two
experiments, participants were shown a videotape of an alleged
psychic bending a key using apparent psychokinetic ability but in
fact using sleight of hand techniques. Participants in one condi-
tion heard the psychic suggest that the key continued to bendafter
being put down on a table, whilst those in a second condition did
not. The findings revealed that those in the suggestion condition
were significantly more likely to report that the key had indeed
continued to bend (even though it had not). The size of this
effect was considerable, with around 40% of the participants
in the suggestion condition reporting that the key continued
to bend compared to virtually no one in the no-suggestion
condition. Somewhat surprisingly, in the light of findings from
the séance studies, no differences were found between believers
in the paranormal and disbelievers in either experiment. In the
second experiment (but not the first), those who erroneously
reported that the key continued to bend were more confident
regarding their recall than those who correctly reported that it
did not. Interestingly, they were also significantly less likely to
remember hearing the actual verbal suggestion provided by the
fake psychic.
Recent studies have applied memory conformity paradigms
to the study of OPEs on the assumption that witnesses of such
events are very likely to discuss what they saw and one person’s
report may influence the memory of other witnesses. Thus, if one
witness to a séance, for example, was initially unsure whether
a particular object had or had not moved during the séance,
confident testimony from a fellow witness that it did may be
sufficient to alter the first witness’s report of the event. Wilson
(2006) used the same basic paradigm as that used by Gabbert
et al. (2003) but with videotapes of two 2.5-min clips of a pseudo-
psychic demonstration of apparently psychokinetic ability. Both
clips contained essentially the same sequence of events but each
included one important piece of information missing in the other
clip, information that gave an indication of how the effect was
achieved. In the first clip, for example, a fork used in a fork-
bending demonstration is clearly handled by the alleged psychic
and in the second, the fork clearly goes out of view. As with the
previous study, the focus of interest was the degree to which the
participants’ recall was distorted as a result of discussion with a
co-witness. Once again it was found that a substantial majority
of participants included crucial items of information about the
event they witnessed that were most likely to have been acquired
as a result of such discussions. This study therefore demonstrated
that, as predicted, memory conformity effects do in fact occur in
apparently paranormal contexts.
This general line of research is important for two main reasons.
The first is that it provides an explanation of reports of various
OPEs in terms of known psychological factors. Opinion polls
repeatedly show that a large proportion of the population believes
in the paranormal and a sizeable minority claims to have had
direct personal experience of paranormal events. But, with a few
notable exceptions, psychology has had little to say about the
origins of such beliefs and experiences until fairly recently. We
strongly believe, in line with other researchers within anomalis-
tic psychology, that it is not enough simply to speculate upon
the various psychological factors that may underlie reports of
OPEs. It is important to support such accounts with empirical
evidence and the current research is aimed at doing precisely
Frontiers in Psychology | Theoretical and Philosophical Psychology November 2014 | Volume 5 | Article 1289 |2
Wilson and French Magic and memory
this with respect to the factors of verbal suggestion and memory
conformity.
The second main reason for carrying out such research is for
what it can tell us about memory more generally. For example,
most previous research into the reliability of eyewitness testimony
has been carried out in a forensic context, often involving the use
of staged crimes and so on. Apart from the obvious importance
in terms of generalisability of studying such effects in a different
context, investigating the reliability of reports of OPEs under
controlled conditions offers an ideal opportunity to demonstrate
the effects of pre-existing beliefs upon perception and memory.
By their very nature, OPEs are often inherently ambiguous and it
is precisely in such circumstances that we would expect top-down
influences upon perception (French, 2001) and memory (French,
2003; French and Wilson, 2006) to be most pronounced.
The choice of belief in the paranormal as a means of exploring
the influence of top-down processes on cognition is particularly
appropriate for several reasons. In addition to the inherent ambi-
guity of most OPEs, (i) paranormal belief is prevalent in all
societies, (ii) belief in the paranormal is very important in many
people’s lives and such beliefs have strong emotional ties (e.g., the
belief in life after death) and (iii) paranormal beliefs often form
part of a larger set of beliefs and attitudes toward such things
as religion, science and indeed mankind’s place in the universe.
Furthermore, standard scales are available to measure the level
of paranormal belief making it an ideal choice for this type of
investigation.
The current study aimed to replicate and extend previous stud-
ies of verbal suggestion and memory conformity by systematically
manipulating both the presence or absence of a verbal suggestion
as well as the type of social influence exerted by a co-witness.
Replication of such effects is crucially important in light of current
concerns regarding poor replicability within psychology (see, e.g.,
Pashler and Wagenmakers, 2012; Ritchie et al., 2012). The study
is based upon Wiseman and Greening’s (2005) demonstration of
the power of verbal suggestion in the context of an alleged demon-
stration of psychokinetic metal-bending. Using the same video
clip as that used in the original study, participants viewed a fake
psychic apparently using psychokinesis to bend a key. After the
psychic had put the bent key down, half of the participants heard
the fake psychic suggest that the key continued to bend while
the other half did not hear the suggestion. It was hypothesized,
in line with the findings of the original study, that those in the
suggestion condition would be more inclined to report that the
key continued to bend in comparison to participants in the no-
suggestion condition.
Furthermore, each participant was also exposed to one of
three types of social influence from a co-witness. One-third of
the participants were exposed to a “negative” social influence,
insofar as the co-witness, during a post-event discussion, reported
that the key did not continue to bend. Another third of the
participants were not exposed to any social influence, as they
did not discuss the demonstration at all. The final third of the
participants were exposed to a “positive” social influence, in that
the co-witness, during the post-event discussion, reported that the
key did indeed continue to bend. The co-witness in the negative
and positive social influence conditions was in fact a stooge. It was
hypothesized, based upon previous memory conformity research,
that the genuine participants in the positive social influence con-
dition would be more inclined to report that the key continued
to bend than those in the no social influence condition, whereas
those in the negative social influence condition would be relatively
less inclined.
Even though Wiseman and Greening (2005) did not find any
difference between believers in the paranormal and disbelievers
in terms of tendency to report that the key continued to bend,
it was hypothesized in the current study that the former group
may show this tendency more strongly on the basis of previous
research including studies of susceptibility to suggestion in the
séance room.
A number of individual difference measures have been shown
to be correlated with both paranormal belief and tendency to
report anomalous experiences on the one hand and susceptibility
to various kinds of memory distortion on the other, including sus-
ceptibility to false memories (French, 2003; French and Wilson,
2006). This suggests that at least some reports of anomalous
events may be based upon false memories. Dissociativity, for
example, has been shown in a number of studies to correlate with
paranormal belief (e.g., Irwin, 1994; Pekala et al., 1995; Wolfradt,
1997; Makasovski and Irwin, 1999; Rattet and Bursik, 2001) and
the tendency to report a wide range of paranormal and anomalous
experiences (e.g., Richards, 1991; Ross et al., 1991; Ross and Joshi,
1992; Pekala et al., 1995), as well susceptibility to false memo-
ries (e.g., Eisen and Carlson, 1998; Hyman and Billings, 1998;
Winograd et al., 1998; Heaps and Nash, 1999; Ost et al., 2005;
Wilson and French, 2006). One possible explanation for the link
between dissociativity and susceptibility to false memories is that,
by definition, high scorers on measures of dissociativity experi-
ence more disruptions in the integration of thoughts, awareness,
and memory. Such individuals may therefore be more prone to
accepting externally presented information as autobiographical
memories.
The relationship between dissociativity and suggestibility is
complex, but several studies have reported a significant correla-
tion using a variety of measures of suggestibility (see Eisen and
Lynn, 2001; Eisen et al., 2002). Therefore, given the known cor-
relation between paranormal belief and dissociativity, a measure
of dissociativity (the Dissociative Experiences Scale, DES) was
administered in order to allow the assessment of possible effects
of dissociativity upon the dependent variables in this study.
Compliance (or eagerness to please) has also been shown to
be related to susceptibility to false memories (e.g., Ost et al.,
2002, 2005). It might be expected that in the current experiment,
where, depending upon the allocated condition, participants may
be exposed to social influence in terms of the initial verbal
suggestion from the fake psychic and/or the comments of the
stooge, level of compliance would be related to the degree to which
participants report that the key continued to bend. Therefore, the
current study also measured compliance, using Snyder’s (1974)
Self-Monitoring Scale (SMS).
The current study complied with the ethical guidelines of the
British Psychological Society and ethical approval to conduct the
study was granted by the Ethical Committee of the Department of
Psychology, Goldsmiths College, University of London.
www.frontiersin.org November 2014 | Volume 5 | Article 1289 |3
Wilson and French Magic and memory
MATERIALS AND METHODS
PARTICIPANTS
One hundred and eighty undergraduates and college employees
from Goldsmiths College, University of London, took part in the
study. Participants were 144 females and 36 males with a mean
age of 24.41 years (SD = 3.45) and an age range of 18–57 years.
All participants responded to a poster advertising for involvement
in an experiment where participants would be asked to judge the
paranormal abilities of a professed psychic. Participants received
either course credit or £5 for their involvement.
DESIGN
This study generally employed a 2 ×3×2 factorial design with
Verbal Suggestion (suggestion vs. no suggestion), Social Influence
(negative social influence vs. no social influence vs. positive
social influence), and Belief Group (believers vs. non-believers),
as between-group factors. The primary dependent variable was
scores on item 3 of a Fixed Response Questionnaire (FRQ3) asking
participants to rate their degree of agreement with the statement
After the key was placed on the table, it continued to bend” (see
below for details).
MATERIALS
Videotape
The videotape used in the study was supplied by Richard Wise-
man and is the same videotape as that used in Wiseman and
Greening’s (2005) experiments. Two versions of the tape were
used. In the suggestion version of the tape the film consists of
a 2-min clip of an interviewer and “psychic” sat at a table with
several objects such as cutlery and keys in front of them. The
interviewer briefly introduces the psychic and invites him to
perform a demonstration of his powers using any of the objects
of his choice. The psychic then picks up a key and appears to
use his psychokinetic powers to bend the key to a 25angle, in
fact achieving this effect by the use of sleight of hand. He then
places the key back on the table and suggests that the key is in
fact still bending, even though it is not. The no verbal suggestion
version of the tape is identical to the suggestion version but part
of the soundtrack was removed so that participants did not hear
the verbal suggestion. The fake psychic used in the demonstration
was in fact a magician who had worked professionally for many
years using sleight of hand techniques.
Questionnaires
Fixed Response Questionnaire. This is the 4-item questionnaire
used by Wiseman and Greening (2005) and consists of statements
concerning the film. Two of the statements are filler items, e.g.,
“The interviewer touched the items on the table.” Responses to
the third item (FRQ3) were used as the main dependent variable
in the study: “After the key was placed on the table, it continued to
bend.” The fourth item on the questionnaire asked participants to
what extent they considered the demonstration involved paranor-
mal forces. For each item, participants were asked to provide their
response on a 7-point scale from 1 (Definitely No) to 7 (Definitely
Yes). Participants were also asked to rate their confidence in their
answers on a similar scale from 1 (not at all confident) to 7 (very
confident).
Forced-choice version of the Australian Sheep-Goat Scale. This
is a widely used scale that consists of 18 statements relating to the
three core concepts of parapsychology: extrasensory perception,
psychokinesis, and life after death. The statements refer to belief
in and alleged experience of the paranormal and respondents are
awarded no points for a “false” response, one point for a “don’t
know” response, and two points for a “true” response (allowing
for a maximum score of 36). Note that this scale was preferred to
the unstandardized Belief in the Paranormal Questionnaire used
by Wiseman and Greening (2005) because it has known validity
and reliability (e.g., Thalbourne and Delin, 1993; Thalbourne,
1995, 2010) and it allowed comparison with other research in this
area (e.g., Wilson and French, 2006). Scores on this scale were
used to allocate participants to belief groups.
Dissociative Experiences Scale. This scale, designed and devel-
oped by Bernstein and Putnam (1986), consists of a 28-item self-
report questionnaire. A typical example would be: “Some people
have the experience of finding new things among their belongings
that they do not remember buying.” Respondents are asked to
circle a box to indicate what percentage of the time this event
happens to them, ranging from 0 to 100% at 10% intervals. Each
item is awarded a score between 0 and 100 and the mean score is
then calculated across the 28 items. The scale has been shown to
have good psychometric properties (Dubester and Braun, 1995)
and internal consistency (Norton et al., 1990).
Self-Monitoring Scale of Expressive Behaviour. This scale, devel-
oped by Snyder (1974), is a 25-item true–false questionnaire
consisting of items such as “When I am uncertain how to act in
a social situation, I look to the behavior of others for cues” and
“My behavior is usually an expression of my true inner feelings,
attitudes, and beliefs.” The score on this scale indicates the extent
to which respondents rely on cues from others in deciding how
to behave in social situations as opposed to relying upon personal
values. One point is awarded for every response in line with such
tendencies.
PROCEDURE
All participants were told that they were to judge the paranor-
mal powers of a professed “psychic” who had claimed to the
Psychology Department that he could demonstrate psychokinetic
ability. Participants were allocated to one of the six experimen-
tal conditions produced by crossing the two factors of Verbal
Suggestion (suggestion vs. no suggestion) and Social Influence
(negative social influence vs. no social influence vs. positive social
influence). To maintain comparability across all experimental
conditions, all participants were tested in pairs. In the positive
and negative social influence conditions, one of the apparent
participants was in fact a stooge playing the part of a co-witness,
whereas in the no social influence condition both participants
were genuine. Participants in the suggestion condition watched
the video with an audible commentary throughout, whereas those
in the no suggestion condition were not presented with the verbal
suggestion from the fake psychic.
In the no social influence conditions, both participants watched
the video and were then asked to complete the questionnaires. In
Frontiers in Psychology | Theoretical and Philosophical Psychology November 2014 | Volume 5 | Article 1289 |4
Wilson and French Magic and memory
the positive and negative social influence conditions, the stooge and
the participant arrived at the testing room at the same time. Both
watched the video but were then told to discuss the details of the
film together. In order to facilitate this discussion, the stooge and
the real participant were asked to complete a short questionnaire
consisting of four questions relating to the film. These included
three filler questions, e.g., “What was the psychic wearing?” and
the crucial question, i.e., “Did the key continue to bend after it had
been placed on the table?” Participants were told to complete this
brief questionnaire together. The stooge was instructed to speak
first and to lead the discussion, either maintaining that the key had
continued to bend and that paranormal forces had been at work
(in the positive condition) or that the key had not continued to
bend and that no paranormal forces were involved (in the negative
condition). After the discussion the participants independently
completed the other questionnaires.
At the end of the experiment, all participants were debriefed
fully. Participants in the positive and negative social influence
conditions were asked if at any time they had suspected that their
fellow co-witness was a confederate of the researcher. However,
no participants reported that they had been suspicious of the
stooge. To maintain continuity the same stooge took part in all
the trials.
RESULTS
Participants were first classified into Belief Groups on the basis
of a median split of Australian Sheep-Goat Scale (ASGS) scores,
with those scoring more than 10 classified as believers and the
rest as disbelievers in the paranormal. It is common practice in
studies comparing high and low paranormal belief groups on per-
formance measures to divide the groups using a median split on
the belief measure as done by Wiseman and Greening (2005) and
in the current study. Although this approach runs the potential
risk of failing to detect real effects because information is lost by
converting a continuous variable to a binary variable (MacCallum
et al., 2002), one can be certain that any effects identified with this
approach would also be found using alternative methods such as
multiple regression. Indeed, results from the current study were
also analyzed using multiple regression techniques and the pattern
of results found was identical to that reported below. However, it
was felt that the effects found were described more clearly using
the results of ANOVAs.
In order to check that unintended sampling bias had not been
introduced by splitting our sample in this way, three 2 ×2×
3 ANOVAs were carried out on the scores from the ASGS, DES,
and SMS, respectively, each with Belief Group, Verbal Suggestion,
and Social Influence as between-group factors. As would be
expected given the method of allocation to belief groups, ASGS
scores were significantly higher for believers (mean = 18.25, SD =
4.80) than disbelievers [mean = 3.78, SD = 3.05; F(1,168) =
562.26, p<0.001]. Also, as expected given the known correlation
between paranormal belief and dissociativity, DES scores were
significantly higher for believers (mean = 36.99, SD = 15.44)
than for disbelievers [mean = 28.05, SD = 16.12; F(1,168) =
12.81, p<0.001]. Interestingly, SMS scores were also significantly
higher for believers (mean = 12.51, SD = 3.65) than disbelievers
[mean = 10.87, SD = 4.08; F(1,168) = 6.04, p= 0.015]. No other
main effects or interactions from any of the three ANOVAs were
statistically significant. ASGS scores correlated significantly with
both DES scores (r= 0.278, p<0.001) and SMS scores (r= 0.180,
p= 0.016) across the sample as a whole.
Next, responses to FRQ3 were analyzed using a 2 ×2×3
ANOVA with the same factors as those used in the previous
analysis. This analysis revealed a significant main effect of Verbal
Suggestion, with participants who heard the suggestion giving
higher ratings on FRQ3 (mean = 3.92, SD = 2.02) than those who
did not [mean = 2.56, SD = 1.74; F(1,168) = 32.40, p<0.001].
A main effect of Social Influence was also found [F(2,168) =
22.01, p<0.001]. Using Bonferroni-adjusted t-tests, it was shown
that positive social influence produced higher ratings on FRQ3
(mean = 4.43, SD = 1.96) than either negative social influence
[mean = 2.50, SD = 1.54; t(118) = 6.02, p<0.001] or no social
influence [mean = 2.78, SD = 1.94; t(118) = 4.63, p<0.001].
However, the two latter conditions did not produce significantly
different ratings [t(118) = 0.89, n.s.]. Finally, believers in the
paranormal gave significantly higher ratings on FRQ3 (mean =
3.75, SD = 1.97) than disbelievers [mean = 2.75, SD = 1.93;
F(1,168) = 9.94, p= 0.002]. No significant interactions were
found.
In light of the significant differences between belief groups
on scores for the DES and SMS, these variables were entered as
covariates in the main analysis of responses to FRQ3 in order
to ascertain whether any differences found between belief groups
could be accounted for in terms of differences between the groups
on these variables. Therefore responses to FRQ3 were analyzed
using a 2 ×2×3 ANOVA with the same factors as those used
in the previous analysis, but with the inclusion of DES and SMS
scores as covariates. This analysis revealed a significant main effect
of Verbal Suggestion, with participants who heard the suggestion
giving higher ratings on FRQ3 (mean = 3.92, SD = 2.02) than
those who did not [mean = 2.56, SD = 1.74; F(1,179) = 32.05,
p<0.001].
A main effect of Social Influence was also found
[F(2,179) = 21.06, p<0.001]. Using Bonferroni-adjusted
t-tests, it was shown that positive social influence produced
higher ratings on FRQ3 (mean = 4.43, SD = 1.96) than either
negative social influence [mean = 2.50, SD = 1.54; t(118) = 6.02,
p<0.001] or no social influence [mean = 2.78, SD = 1.94;
t(118) = 4.63, p<0.001]. However, the two latter conditions did
not produce significantly different ratings [t(118) = 0.89, n.s.].
Finally, even with DES and SMS scores entered as covariates,
believers in the paranormal gave significantly higher ratings on
FRQ3 (mean = 3.75, SD = 1.97) than disbelievers [mean = 2.75,
SD = 1.93; F(1,179) = 7.89, p= 0.006]. DES and SMS scores were
not significantly related to responses on the FRQ3 in this analysis.
Once again, no significant interactions were found.
Following Wiseman and Greening (2005), participants were
then allocated to two groups depending upon their responses
to the FRQ3. Those who responded with either a 5, 6, or 7
were allocated to the key continued to bend group. The rest were
allocated to the key did not continue to bend group. The numbers
and percentages in each group across experimental conditions
are presented in Table 1. Chi-square analyses between group
and suggestion within each social influence condition revealed
www.frontiersin.org November 2014 | Volume 5 | Article 1289 |5
Wilson and French Magic and memory
Table 1 | Numbers and percentages of participants in the key
continued to bend and the key did not continue to bend groups
across experimental conditions.
Key continued
to bend group
Key did not continue
to bend group
Negative social influence
Suggestion 7 (23.3%) 23 (76.7%)
No suggestion 2 (6.7%) 28 (93.3%)
No social influence
Suggestion 10 (33.3%) 20 (67.7%)
No suggestion 0 (0%) 30 (100%)
Positive social influence
Suggestion 18 (60%) 12 (40%)
No suggestion 12 (40%) 18 (60%)
Table 2 | Mean confidence ratings (SDs in parentheses) given to item
FRQ3 by participants in the key continued to bend and the key did
not continue to bend groups across experimental conditions.
Key continued
to bend group
Key did not continue
to bend group
Negative social influence 4.56 (1.59), N= 9 5.53 (1.84), N= 51
No social influence 6.20 (1.48), N= 10 5.24 (1.62), N= 50
Positive social influence 6.07 (1.02), N= 30 4.73 (1.82), N= 30
a highly significant effect (χ2= 12.0, df = 1, p= 0.001), in
the no social influence condition (thus replicating Wiseman and
Greening, 2005) with 10 participants (33.3%) reporting that the
key continued to bend if given the verbal suggestion compared
to none in the no-suggestion condition. Neither of the other
chi-square analyses was significant. It is worth noting that the
percentage of participants in the suggestion condition reporting
that the key continued to bend was decreased to 23.3% in the
negative social influence condition and almost doubled to 60%
in the positive social influence condition (χ2= 6.9, df = 1,
p= 0.009).
Wiseman and Greening (2005, Experiment 2) found that those
reporting that the key continued to bend were more confident
about the accuracy of their report than those who reported that
it did not (although this result was not found in their first exper-
iment). Responses to item 3b of the FRQ in the current study,
indicating confidence in the accuracy of participants’ reports on
FRQ3, are presented in Table 2. These data were subjected to a
2×3 ANOVA with Bend Group (did continue to bend vs. did
not continue to bend) and Social Influence Group as between-
groups factors. No significant main effects were found but a highly
significant interaction was revealed [F(2,179) = 5.25, p= 0.006].
Further exploration of this interaction, using three Bonferroni-
adjusted t-tests, revealed only one significant effect: participants
in the positive social influence condition who reported that the
key continued to bend were far more confident in their ratings
than those who reported that the key did not continue to bend
[t(58) = 3.51, p= 0.001]. The same general trend was evident
for those in the no social influence group although the oppo-
site trend was evident for those in the negative social influence
group, i.e., in the latter condition, those reporting that the key
Table 3 | Number and percentages of participants in the
demonstration was paranormal and the demonstration was not
paranormal groups across the experiment as a whole, broken down
by Belief Group and Bend Group.
Demonstration
was paranormal
Demonstration was
not paranormal
Believers
Key continued to bend 13 (39.4%) 20 (60.6%)
Key did not continue to bend 7 (12.7%) 48 (87.3%)
Disbelievers
Key continued to bend 1 (6.2%) 15 (93.8%)
Key did not continue to bend 2 (2.6%) 74 (97.4%)
continued to bend were relatively less confident than those who
reported that it did not. A 2 ×2 ANOVA using the same fac-
tors as those in the previous analysis but excluding the positive
social influence group revealed that the interaction between Bend
Group and Social Influence was still significant [F(1,119) = 5.13,
p= 0.025].
Wiseman and Greening (2005) did not report any analyses
of responses from the fourth item on the FRQ (FRQ4), deal-
ing with the degree to which participants believed the demon-
stration, including the initial key bending by sleight of hand,
involved paranormal forces. FRQ4 data from the present study
were subjected to a 2 ×2×3 ANOVA with Belief Group,
Verbal Suggestion, and Social Influence as between-group fac-
tors. Not surprisingly, believers in the paranormal gave higher
ratings (mean = 3.32, SD = 1.64) than disbelievers [mean = 1.87,
SD = 1.18; F(1,179) = 44.38, p<0.001]. Perhaps more sur-
prisingly, higher ratings were given by participants exposed to
the verbal suggestion (mean = 2.90, SD = 1.73) than those who
were not so exposed [mean = 2.26, SD = 1.39; F(1,179) = 9.57,
p= 0.002]. The generally low levels of ratings of paranormality
should, however, be noted.
Participants were then allocated to groups on the basis of
whether they did or did not believe the demonstration involved
paranormal forces. Those scoring either 5, 6, or 7 on FRQ4 were
allocated to the demonstration was paranormal group and the rest
were allocated to the demonstration was not paranormal group.
The numbers and percentages in each group across the experi-
ment as a whole are presented in Table 3. Across the experiment as
a whole, 49 out of 180 participants (27.2%) reported that the key
continued to bend and 23 (12.8%) believed they had witnessed
something paranormal in the demonstration as a whole. Of the
88 believers, 33 (37.5%) reported that the key continued to bend
and 20 (22.7%) believed they had witnessed paranormal forces
in action. Note that this implies that many of the believers who
reported that the key carried on bending did not believe that this
particular demonstration involved genuine paranormal forces,
presumably believing instead that it was based upon some form
of trickery. Of the 92 disbelievers, only 16 (17.4%) reported that
the key continued to bend and only 3 (3.2%) of those classified as
disbelievers reported that they had witnessed paranormal forces
in action. Presumably, this tiny percentage of “disbelievers” who
believed they had witnessed a genuine paranormal event had been
so classified because they did not believe in life after death and/or
Frontiers in Psychology | Theoretical and Philosophical Psychology November 2014 | Volume 5 | Article 1289 |6
Wilson and French Magic and memory
ESP even though, evidently, they did believe in psychokinesis. It
is interesting to compare the interpretation of the demonstration
between the belief groups for those participants who reported
that the key did continue to bend. Of 33 believers who reported
that the key continued to bend, 13 (39.4%) reported that the
demonstration was paranormal. Of 16 disbelievers who reported
that the key continued to bend, only one (6.2%) reported that
the demonstration was paranormal (χ2= 5.8, df = 1, p= 0.016).
Clearly, disbelievers were much more likely to opt for a non-
paranormal explanation even if they believed they had seen the
key carry on bending.
DISCUSSION
The results of this experiment largely confirm the basic finding
of Wiseman and Greening’s (2005) experiments; that is, in this
context, a relatively mild verbal suggestion from a fake psychic
that a bent key continued to bend after it had been placed upon
a table was sufficient to lead a substantial number of witnesses
to erroneously report that the key had indeed done just that. In
the no social influence condition in the current experiment, the
condition most similar to that used by Wiseman and Greening
(2005), one-third of the participants reported continued bending,
compared to 39.13% in their Experiment 1 and 36.54% in their
Experiment 2. Also in line with Wiseman and Greening’s (2005)
findings, no participants in the no suggestion and no social influ-
ence condition reported continued bending.
The current study extended the findings of the original experi-
ments by incorporating an additional social influence component
into the design. When a stooge co-witness insisted that the key
continued to bend, 60% of the participants agreed. When the
stooge co-witness insisted that the key did not continue to bend,
the percentage who reported that it did was substantially reduced,
but even then 23.3% reported that it did. This is a powerful
demonstration that it is not only what witnesses to an ostensibly
paranormal event believe that they have actually perceived at
the time that determines their subsequent reports but that such
reports will also be influenced by discussion with co-witnesses in
line with findings from memory conformity research.
We also found one result that was not in line with the findings
of the original experiments by Wiseman and Greening (2005).
In the current study, believers in the paranormal were found to
be more likely to report that the key continued to bend com-
pared to disbelievers. Wiseman and Greening (2005) considered
two possible explanations for their failure to find any difference
between belief groups. First, they considered the possibility that
previous studies reporting an association between paranormal
belief and suggestibility might be mistaken, possibly reflecting a
“file-drawer” effect in which a few studies finding a spuriously
significant relationship between these two variables had been
published but that they should be considered in the wider context
of a possibly much larger number of studies that had tried and
failed to find such an effect and had therefore never been sub-
mitted for publication. Second, they suggested that paranormal
belief may correlate with certain kinds of suggestibility but not
the form of suggestibility involved in their key-bending experi-
ments. The current findings would argue against both of these
suggestions. It appears that the type of suggestibility involved in
both the original experiments and the current study is indeed
correlated with paranormal belief. The most likely explanation
for the discrepancy between Wiseman and Greening’s (2005)
findings in this regard and the findings of the current study is
our decision to use the ASGS as a measure of paranormal belief.
Furthermore, the belief-related effects found were not explicable
in terms of differences between the belief groups on the DES and
SMS measures.
Wiseman et al. (2003), in the context of discussing the effects
of suggestion on eyewitness reports in the séance room, acknowl-
edge that it is often difficult to determine whether verbal sug-
gestion directly affects the perception of the event, memory for
the event, or both. It is even possible that neither is affected
and that the results are due to demand characteristics. Thus, it
is possible that the verbal suggestions during the séance directly
influence the perception of the witnesses in such a way that those
witnesses who are exposed to such suggestions actually perceive
stationary objects to be moving in real time. Alternatively, it is
possible that the witnesses did not actually perceive the stationary
objects to be moving at the time but that their memories of
the event were affected by the verbal suggestions when, 2 weeks
later, they received a questionnaire asking them to recall details
of the séance. By that time, their memory for the séance would
be beginning to fade and, in their attempts to reconstruct the
details of what happened, they may have blended the fake psy-
chic’s suggestions in with their blurred memory of the original
event in such a way that they now recalled stationary objects
as moving. Finally, it is possible that at the time the partici-
pants completed the recall questionnaire, they did not actually
believe that the stationary objects had moved at all, but simply
reported that they did, perhaps believing that this would please
the investigators.
The results of the current study can perhaps cast some light
upon these competing explanations. We begin by acknowledg-
ing that self-report data alone can never definitively distinguish
between perceptual effects and memory effects. Even if we ask
participants to tell us what they are perceiving as events unfold
before them, there will always be a slight delay, perhaps only
a fraction of a second, between the perception of the events
and the subsequent report. Thus it is always possible to argue
that the perception of the events was fundamentally veridical
but the memory of the event was somehow distorted. In fact,
however, the general position of modern cognitive psychology
is that perception and memory are constructive processes and
that both will be influenced by bottom-up influences (i.e., raw
sensory input) and top-down influences (e.g., beliefs, knowledge,
expectations). Thus perception itself is heavily dependent upon
memory. When considering ostensibly paranormal events then,
both perception (French, 2001) and memory (French, 2003) are
likely to be influenced by a variety of top-down influences and
thus both perception and memory are likely to be influenced by
verbal suggestions that alter expectations.
In considering the experimental set-up used by Wiseman
and Greening (2005), it appears that participants completed the
FRQ immediately after viewing the video, thus minimizing the
possibility that the effect is due to the type of blending of a
blurred memory of an essentially accurate perception with the
www.frontiersin.org November 2014 | Volume 5 | Article 1289 |7
Wilson and French Magic and memory
memory of the suggestion, as described above. Although a direct
effect upon the actual perception of the event is entirely consistent
with Wiseman and Greening’s (2005) data, the possibility of an
explanation in terms of demand characteristics remains.
The data from the current study demonstrate unequivocally
that social influence provided after the video had been viewed was
sufficient to alter witnesses’ reports of what they saw. Fully 40% of
the participants in the positive social influence condition reported
that the key continued to bend even in the absence of a verbal
suggestion to that effect from the fake psychic. Furthermore,
of those participants who did receive the verbal suggestion, the
percentage of participants reporting that the key continued to
bend was markedly affected by the reports of the stooge co-
witness. Such effects are only explicable as either memory effects
or in terms of demand characteristics.
We do not feel that demand characteristics provide a par-
simonious explanation of our findings when the responses to
item FRQ3b are considered. Data relating to the confidence
expressed in the memory report indicate that in both the no
social influence and the positive social influence conditions, par-
ticipants erroneously reporting that the key continued to bend
expressed higher levels of confidence than those who did not
report that the key continued to bend, thus replicating Wiseman
and Greening’s (2005) Experiment 2. In both cases, confidence
levels were extremely high (>6 on a 7-point scale). This clearly
indicates that expressions of confidence in the accuracy of reports
of OPEs should not be taken as any kind of indication of reli-
ability. The lowest confidence ratings in the experiment came
from those participants in the negative social influence condition
who reported that the key continued to bend and those in the
positive social influence condition who reported that it did not.
It seems likely that the former group really did perceive the key
as continuing to bend and were prepared to stick to that view
despite a forceful stooge arguing that it did not. The latter group,
on the other hand, did not report that the key continued to
bend but their confidence in that view was clearly shaken by a
forceful stooge arguing that it had. In both cases, the responses of
participants seem to be more in line with participants trying their
best to give honest accounts of what they saw rather than behaving
in accordance with demand characteristics.
Finally, there is another level at which the influence of beliefs
comes into play. The overall interpretation of the demonstration
as evidence for the paranormal was, as one might expect, strongly
related to paranormal belief. Considering first the believers, it is
worth noting that the vast majority did not consider that the
demonstration involved paranormal forces—even if they reported
that the key continued to bend. Even so, a much higher pro-
portion of believers than disbelievers reported that they had
witnessed paranormal forces in action (around 40% of those
who reported continued bending of the key). The disbelievers,
on the other hand, were much less likely to report that the key
continued to bend and, even if they thought it did, they were
much less likely to opt for a paranormal explanation. Across the
experiment as a whole, of 49 participants who reported that the
key continued to bend, only 14 thought that the demonstration
involved paranormal forces. The others, presumably, thought that
it was some kind of trick, a tendency found much more strongly
amongst the disbelievers than amongst the believers. This is not
unreasonable, given that such an effect could have been produced
by either special effects or by the use of a trick key. It is even
possible that some participants realized it was a simple effect of
suggestion but were honest enough to admit that it had worked
on them. It would be of interest in future studies to ask such
participants directly for their explanation of the effect. It would
also be of interest in future investigations to include conditions in
which the key really does appear to bend to investigate whether
disbelievers are prone to deny such events.
It should be noted that one difference between the two social
influence conditions and the no social influence condition in the
current study was that the former involved discussion of what had
been witnessed whereas the latter did not. It is therefore possible
that this might have influenced the results in some way, e.g., in
terms of differential delay in recall, differences in the number of
retrieval attempts, etc. We do not feel this was a major method-
ological problem with our study as these factors were matched
across the positive and negative social influence conditions and
thus the impact of different types of social influence are clearly
demonstrated by our results. However, we would recommend that
similar studies in future replace our current no social influence
condition with one that does involve discussion with the stooge
participant and the use of the short (4-question) questionnaire
ostensibly to facilitate the discussion. The difference would be
that the stooge would be presented as someone who has not
themselves seen the video clip and ostensibly is simply acting as
a facilitator.
ACKNOWLEDGMENTS
Thanks are due to Richard Wiseman for providing the video clip
used in this study and to Andy Marriott for playing the part of the
stooge.
REFERENCES
Bernstein, E. M., and Putnam, F. W. (1986). Development, reliability, and validity
of a dissociation scale. J. Nerv. Ment. Dis. 174, 727–735. doi: 10.1097/00005053-
198612000-00004
Besterman, T. (1932). The psychology of testimony in relation to paraphysical
phenomena: report of an experiment. Proc. Soc. Psych. Res. 40, 363–387.
Dubester, K. A., and Braun, B. G. (1995). Psychometric properties of the Dissocia-
tive Experiences Scale. J. Nerv. Ment. Dis. 183, 231–235. doi: 10.1097/00005053-
199504000-00008
Eakin, D. K., Schreiber, T. A., and Sergent-Marshall, S. (2003). Misinformation
effects in eyewitness memory: the presence and absence of memory impairment
as a function of warning and misinformation accessibility. J. Exp. Psychol. Learn.
Mem. Cogn. 29, 813–825. doi: 10.1037/0278-7393.29.5.813
Eisen, M. L., and Carlson, E. B. (1998). Individual differences in suggestibil-
ity: examining the influence of dissociation, absorption, and a history of
childhood abuse. Appl. Cogn. Psychol. 12, S47–S61. doi: 10.1002/(SICI)1099-
0720(199812)12:7<S47::AID-ACP598>3.0.CO;2-P
Eisen, M. L., and Lynn, S. J. (2001). Dissociation, memory and suggestibility in
adults and children. Appl. Cogn. Psychol. 15, S49–S73. doi: 10.1002/acp.834
Eisen, M. L., Morgan, D. Y., and Mickes, L. (2002). Individual differences in eye-
witness memory and suggestibility: examining relations between acquiescence,
dissociation and resistance to misleading information. Pers. Individ. Differ. 33,
553–571. doi: 10.1016/S0191-8869(01)00172-6
French, C. C. (1992). Factors underlying belief in the paranormal: do sheep and
goats think differently? Psychologist 5, 295–299.
French, C. C. (2001). Paranormal Perception? A Critical Evaluation, Monograph
Series No. 42. London: Institute for Cultural Research.
Frontiers in Psychology | Theoretical and Philosophical Psychology November 2014 | Volume 5 | Article 1289 |8
Wilson and French Magic and memory
French, C. C. (2003). Fantastic memories: the relevance of research into eyewitness
testimony and false memories for reports of anomalous experiences. J. Con-
scious. Stud. 10, 153–174.
French, C. C., and Stone, A. (2014). Anomalistic Psychology: Exploring Paranormal
Belief and Experience. Basingstoke: Palgrave Macmillan.
French, C. C., and Wilson, K. (2006). Incredible memories: how accurate are
reports of anomalous events? Eur. J. Parapsychol. 21, 166–181.
French, C. C., and Wilson, K. (2007). “Cognitive factors underlying paranormal
beliefs and experiences,” in Tall Tales: Popular Myths About the Mind and Brain,
ed. S. Dalla Sala (Oxford: Oxford University Press), 3–22.
Gabbert, F., and Hope, L. (2013). “Suggestibility and memory conformity,” in
Suggestibility in legal contexts: Psychological research and forensic implications, eds
A. M. Ridley, F. Gabbert, and D. J. La Rooy (London: Wiley), 65–83.
Gabbert, F., Memon, A., and Allan, K. (2003). Memory conformity: can witnesses
influence each other’s memories for an event? Appl. Cogn. Psychol. 17, 533–543.
doi: 10.1002/acp.885
Heaps, C., and Nash, M. (1999). Individual differences in imagination inflation.
Psychon. Bull. Rev. 6, 313–318. doi: 10.3758/BF03214120
Hodgson, R., and Davey, S. J. (1887). The possibilities of mal-observation and
lapse of memory from a practical point of view. Proc. Soc. Psych. Res. 4,
381–495.
Hyman, I. E. Jr., and Billings, F. J. (1998). Individual differencesand the creation of
false childhood memories. Memory 6, 1–20. doi: 10.1080/741941598
Irwin, H. J. (1994). Paranormal belief and proneness to dissociation. Psychol. Rep.
75, 1344–1346. doi: 10.2466/pr0.1994.75.3.1344
Loftus, E. F. (1975). Leading questions and the eyewitness report. Cogn. Psychol. 7,
560–572. doi: 10.1016/0010-0285(75)90023-7
Loftus, E. F., and Pickrell, J. E. (1995). The formation of false memories. Psychiatr.
Ann. 25, 720–725. doi: 10.3928/0048-5713-19951201-07
MacCallum, R. C., Zhang, S., Preacher, K. J., and Rucker, D. D. (2002). On the
practice of dichotomization of quantitative variables. Psychol. Methods 7, 19–40.
doi: 10.1037/1082-989X.7.1.19
Makasovski, T., and Irwin, H. J. (1999). Paranormal belief, dissociative tendencies,
and parental encouragement of imagination in childhood. J. Am. Soc. Psych. Res.
93, 233–247.
Norton, G. R., Ross, C. A., and Novotnig, M. F. (1990). Factors that predict
scores on the Dissociative Experiences Scale. J. Clin. Psychol. 46, 273–277. doi:
10.1002/1097-4679(199005)46:3<273::AID-JCLP2270460304>3.0.CO;2-X
Ost, J., Foster, S., Costall, A., and Bull, R. (2005). False reports of child-
hood events in appropriate interviews. Memory 13, 700–710. doi: 10.1080/
09658210444000340
Ost, J., Vrij,A., Costall, A., and Bull, R. (2002). Crashing memories and reality mon-
itoring: distinguishing between perceptions, imaginations and ‘false memories’.
Appl. Cogn. Psychol. 16, 125–134. doi: 10.1002/acp.779
Pashler, H., and Wagenmakers, E.-J. (2012). Editors’ introduction to the special
section on replicability in psychological science: a crisis in confidence? Perspect.
Psychol. Sci. 7, 528–530. doi: 10.1177/1745691612465253
Pekala, R. J., Kumar, V. K., and Marcano, G. (1995). Anomalous/paranormal
experiences, hypnotic susceptibility, and dissociation. J. Am. Soc. Psych. Res. 89,
313–332.
Rattet, S. L., and Bursik, K. (2001). Investigating the personality correlates of
paranormal belief and precognitive experience. Pers. Individ. Differ.31, 433–444.
doi: 10.1016/S0191-8869(00)00148-3
Richards, D. G. (1991). A study of the correlations between subjective psychic
experiences and dissociative experiences. Dissociation 4, 83–91.
Ritchie, S. J., Wiseman, R., and French, C. C. (2012). Replication, replication,
replication. Psychologist 25, 346–348.
Ross, C. A., and Joshi, S. (1992). Paranormal experiences in the general population.
J. Nerv. Ment. Dis. 180, 357–361. doi: 10.1097/00005053-199206000-00004
Ross, C. A., Ryan, L., Voigt, H., and Eide, L. (1991). High and low dissociators in a
college student population. Dissociation 4, 147–151.
Snyder, M. (1974). The self-monitoring of expressive behavior. J. Pers. Soc. Psychol.
30, 526–537. doi: 10.1037/h0037039
Thalbourne, M. A. (1995). Further studies of the measurement and correlates of
belief in the paranormal. J. Am. Soc. Psych. Res. 89, 233–247.
Thalbourne, M. A. (2010). The Australian Sheep-Goat Scale: development and
empirical findings. Aust. J. Parapsychol. 10, 5–39.
Thalbourne, M. A., and Delin, P. S. (1993). A new instrument for measuring the
sheep-goat variable: its psychometric properties and factor structure. J. Soc.
Psych. Res. 59, 172–186.
Wilson, K. (2006). Memory for Anomalous Events. Unpublished doctoral disserta-
tion, Goldsmiths, University of London, London.
Wilson, K., and French, C. C. (2004). Memory conformity and paranormal belief.
Paper Presented at the 47th Annual Convention of the Parapsychological Associa-
tion, Vienna.
Wilson, K., and French, C. C. (2006). The relationship between susceptibility
to false memories, dissociativity, and paranormal belief and experience. Pers.
Individ. Differ. 41, 1493–1502. doi: 10.1016/j.paid.2006.06.008
Winograd, E., Peluso, J. P., and Glover, T. A. (1998). Individual differences
in susceptibility to memory illusions. Appl. Cogn. Psychol. 12, S5–S27. doi:
10.1002/(SICI)1099-0720(199812)12:7<S5::AID-ACP553>3.0.CO;2-D
Wiseman, R., and Greening, E. (2005). ‘It’s still bending’: verbal sugges-
tion and alleged psychokinetic ability. Br. J. Psychol. 96, 115–127. doi:
10.1348/000712604X15428
Wiseman, R., Greening, E., and Smith, M. (2003). Belief in the paranor-
mal and suggestion in the seance room. Br. J. Psychol. 94, 285–297. doi:
10.1348/000712603767876235
Wiseman, R., and Morris, R. L. (1995). Recalling pseudo-psychic demonstrations.
Br. J. Psychol. 86, 113–125. doi: 10.1111/j.2044-8295.1995.tb02549.x
Wiseman, R., Smith, M., and Wiseman, J. (1995). Eyewitness testimony and the
paranormal. Skeptic. Inq. 19, 29–32.
Wolfradt, U. (1997). Dissociative experiences, trait anxiety and paranormal beliefs.
Pers. Individ. Differ. 23, 15–19. doi: 10.1016/S0191-8869(97)00043-3
Wright, D. B., Memon, A., Skagerberg, E. M., and Gabbert, F. (2009). When
eyewitnesses talk. Curr. Dir. Psychol. Sci. 18, 174–178. doi: 10.1111/j.1467-
8721.2009.01631.x
Wright, D. B., Self, G., and Justice, C. (2000). Memory conformity: exploring
misinformation effects when presented by another person. Br. J. Psychol. 91,
189–202. doi: 10.1348/000712600161781
Conflict of Interest Statement: The Review Editor Fiona Gabbert declares that,
despite being affiliated with the same institution as authors Krissy Wilson and
Christopher French, the review process was handled objectively and no conflict of
interest exists. The authors declare that the research was conducted in the absence
of any commercial or financial relationships that could be construed as a potential
conflict of interest.
Received: 20 August 2014; accepted: 23 October 2014; published online: 13 November
2014.
Citation: Wilson K and French CC (2014) Magic and memory: using conjuring
to explore the effects of suggestion, social influence, and paranormal belief on eye-
witness testimony for an ostensibly paranormal event. Front. Psychol. 5:1289. doi:
10.3389/fpsyg.2014.01289
This article was submitted to Theoretical and Philosophical Psychology, a section of
the journal Frontiers in Psychology.
Copyright © 2014 Wilson and French. This is an open-access article distributed under
the terms of the Creative Commons Attribution License (CC BY). The use, distribution
or reproduction in other forums is permitted, provided the original author(s) or
licensor are credited and that the original publication in this journal is cited, in
accordance with accepted academic practice. No use, distribution or reproduction is
permitted which does not comply with these terms.
www.frontiersin.org November 2014 | Volume 5 | Article 1289 |9
... We wonder if these magic memories are similar to those acquired in especially emotional circumstances (flashbulb memories), in which the vivid memory of the experience does not guarantee the trustworthiness of its details (Hirst et al., 2015). Along these lines, memorability studies on supposedly paranormal experiences (some published more than 130 years ago) show that memories are very unreliable, and that, depending on the circumstances, there is a propensity to remember events that have not happened (Hodgson & Davy, 1887;Besterman, 1932;Wiseman & Morris, 1995;Wilson & French, 2014). Subjects are also susceptible to manipulation through suggestion and instructions (Wiseman, Greening & Smith, 2003;Wiseman & Greening, 2005;Wilson & French, 2014). ...
... Along these lines, memorability studies on supposedly paranormal experiences (some published more than 130 years ago) show that memories are very unreliable, and that, depending on the circumstances, there is a propensity to remember events that have not happened (Hodgson & Davy, 1887;Besterman, 1932;Wiseman & Morris, 1995;Wilson & French, 2014). Subjects are also susceptible to manipulation through suggestion and instructions (Wiseman, Greening & Smith, 2003;Wiseman & Greening, 2005;Wilson & French, 2014). We believe that these studies can be a good reference to plan the necessary and yet non-existent studies on the subsequent memorability of magic shows. ...
Article
Full-text available
Cognitive scientists have paid very little attention to magic as a distinctly human activity capable of creating situations that are considered impossible because they violate expectations and conclude with the apparent transgression of well-established cognitive and natural laws. This illusory experience of the “impossible” entails a very particular cognitive dissonance that is followed by a subjective and complex “magical experience”. Here, from a perspective inspired by visual neuroscience and ecological cognition, we propose a set of seven fundamental cognitive phenomena (from attention and perception to memory and decision-making) plus a previous pre-sensory stage that magicians interfere with during the presentation of their effects. By doing so, and using as an example the deconstruction of a classic trick, we show how magic offers novel and powerful insights to study human cognition. Furthermore, live magic performances afford to do so in tasks that are more ecological and context-dependent than those usually exploited in artificial laboratory settings. We thus believe that some of the mysteries of how the brain works may be trapped in the split realities present in every magic effect.
... We wonder if these magic memories are similar to those acquired in especially emotional circumstances (flashbulb memories), in which the vivid memory of the experience does not guarantee the trustworthiness of its details (Hirst et al. 2015). Along these lines, memorability studies on supposedly paranormal experiences (some published more than 130 years ago) show that memories are very unreliable, and that, depending on the circumstances, there is a propensity to remember events that have not happened (Hodgson and Davey, 1887, Besterman 1932, Wiseman and Morris 1995, Wilson and French 2014. Subjects are also susceptible to manipulation through suggestion and instructions (Wiseman et al. 2003, Wiseman and Greening 2005, Wilson and French 2014. ...
... Along these lines, memorability studies on supposedly paranormal experiences (some published more than 130 years ago) show that memories are very unreliable, and that, depending on the circumstances, there is a propensity to remember events that have not happened (Hodgson and Davey, 1887, Besterman 1932, Wiseman and Morris 1995, Wilson and French 2014. Subjects are also susceptible to manipulation through suggestion and instructions (Wiseman et al. 2003, Wiseman and Greening 2005, Wilson and French 2014. We believe that these studies can be a good reference to plan the necessary and non-existent studies on the subsequent memorability of magic shows, and it is tempting to speculate that only the emotional details of the magic experience are remembered. ...
Preprint
Cognitive scientists have paid very little attention to magic as a distinctly human activity capable of creating situations or events that are considered impossible because they violate expectations and conclude with the apparent transgression of well-established cognitive and natural laws. And even though magic techniques appeal to all known cognitive processes from sensing, attention and perception to memory and decision making, the relation between science and magic has so far been mostly unidirectional, with the primary goal of unraveling how magic works. Building up from the deconstruction of a classic magic trick, we provide here a cognitive foundation for the use of magic as a unique and largely untapped research tool to dissect cognitive processes in tasks arguably more natural than those usually exploited in artificial laboratory settings. Magicians can submerge every spectator into the precise experimental protocol they have previously designed, accounting with ease for both circumstantial and social contexts. Magicians do not base the success of their experiments in statistical measures that smear out the individual in favor of an average spectator that we know never exists in the real world. They target each and everyone in the audience and, often, with a complete accomplishment. Magicians deliver their cognitive manipulations in real-time, in tight closed-loop with the audience, and in a single trial (they cannot afford to repeat the trick if it fails). Magic has also an inherent and strong social component, merging the private cognitive processes of each spectator with the group dynamics. Finally, when combined with the wide range of precise measuring and wearable technologies available today, magic paves the way for a road not taken towards real-world cognitive science. We dare to speculate that some of the mysteries of how the brain works may be trapped in the split realities present in each magic effect.
... , temporal attention (Barnhart, Ehlert, Goldinger, & Mackey, 2018;Rieiro, Martinez-Conde, & Macknik, 2013), and eyewitness memory (Wilson & French, 2014). ...
Article
Full-text available
The methods of magicians provide powerful tools for enhancing the ecological validity of laboratory studies of attention. The current research borrows a technique from magic to explore the relationship between microsaccades and covert attention under near-natural viewing conditions. We monitored participants’ eye movements as they viewed a magic trick where a coin placed beneath a napkin vanishes and reappears beneath another napkin. Many participants fail to see the coin move from one location to the other the first time around, thanks to the magician’s misdirection. However, previous research was unable to distinguish whether or not participants were fooled based on their eye movements. Here, we set out to determine if microsaccades may provide a window into the efficacy of the magician’s misdirection. In a multi-trial setting, participants monitored the location of the coin (which changed positions in half of the trials), while engaging in a delayed match-to-sample task at a different spatial location. Microsaccades onset times varied with task difficulty, and microsaccade directions indexed the locus of covert attention. Our com-bined results indicate that microsaccades may be a useful metric of covert attentional processes in applied and ecologically valid settings.
... At rehearsal signpost eight the walker was urged to observe surroundings and find a metaphor that captured their current life situation. To enhance the potential wellbeing effects, the language used on all of psychological signposts was suggestive and in five signposts contained phrases such as "Feel your mind and body becoming calm" and "Feel your mood improve" (Sinclair et al., 1997;Wilson and French, 2014;Wiseman and Greening, 2005). ...
Article
The purpose of the study was to investigate whether deliberate psychological tasks, intended to focus people’s attention on the interaction between themselves and natural surroundings, are linked with mood enhancement and self-reported restoration. In four European countries (Finland, France, Luxembourg, Sweden), we surveyed the experiences of volunteers (N = 299) who walked forest trails and carried out psychological tasks printed on the signposts along them. We investigated the similarities and differences of the trail experiences between the countries. Via multigroup modeling, we further examined the moderating role of nature-connectedness in relationships between satisfaction with the contents of the psychological tasks, mood enhancement, and restorative benefits. The results showed that, independent of age and gender, participants were more satisfied with the trails in Sweden and Luxembourg than in Finland. We detected no reliable differences in the restorative experiences or willingness to recommend the trail for others. In the moderation model, satisfaction with the signposts’ contents was connected to positive restorative change and mood enhancement. The moderator effects of nature-connectedness were not significant for either outcome. Thus, it is likely that satisfactory tasks will work equally well for people varying in nature-connectedness. This is a promising prospect for public health promotion. The fairly high level of nature-connectedness among the participants limits the generalizability of our results. Conclusions concerning the role of nature-connectedness should be made with caution due to the limited coverage of the concept in our measure. Future studies that separate the effect of psychological tasks from the restorative effects of nature itself are needed. Keywords Engagement; Psychological tasks; Forest trail; Mood; Restoration; Nature-connectedness
... Critics of Plantinga's argument also note that although our cognitive faculties are indeed fallible and can sometimes give rise to false beliefs, Plantinga presents too strong a dichotomy, making either all of our beliefs or none of them reliable (Boudry and Vlerick 2014). In fact, it has long been known that our cognitive faculties are sometimes prone to biases (Nickerson 1998;Schacter, Guerin, and St. Jacques 2011;Wilson and French 2014), but these biases can only be detected against a background of accurate and reliable belief formation. If our biases were as pervasive as Plantinga suggests, scientists could not even investigate them, because they could never be escaped from. ...
Article
Alvin Plantinga's evolutionary argument against naturalism states that evolution cannot produce warranted beliefs. In contrast, according to Plantinga, Christian theism provides (I) properly functioning cognitive faculties in (II) an appropriate cognitive environment, in accordance with (III) a design plan aimed at producing true beliefs. But does theism fulfill criteria I–III? Judging from the Bible, God employs deceit in his relations with humanity, rendering our cognitive functions unreliable (I). Moreover, there is no reason to suppose that God's purpose would be to produce true beliefs in humans (III). Finally, from the theistic/religious perspective, it is impossible to tell whether observations have natural or supernatural causes, which undermines an appropriate cognitive environment (II). Reliable identification of deceit or miracles could alleviate these problems, but the theistic community has failed to resolve this issue. Dismissal of parts of the Bible, or attempts to find alternative interpretations, would collapse into skepticism or deism. Thus, Plantinga's problem of epistemic warrant backfires on theism.
... For example, it could generate false memories in the spectators' minds (e.g., Loftus, 1997;Loftus, 1992; see also Lamont & Wiseman, 2001, for the Indian rope illusion). Wilson and French (2014) replicated Wiseman and Greening's (2005) experiment with an accomplice witness. Results showed that after watching video, when the accomplice witness suggested that the key was still bending, more participants reported that the key continued to bend on the table compared with a condition without witness or a condition with a negative suggestion (i.e., "the key did not continue to bend"). ...
... The reported false memories of broken glass could not have been derived directly from the video, because the video did not actually show any broken glass; thus, the false memory was arguably induced by the question itself. Other researchers have demonstrated that false verbal suggestions presented co-currently with events can also induce false reports (Wiseman et al., 2003;Wiseman and Greening, 2005;Wilson and French, 2014). ...
Article
Drawing inspiration from sleight-of-hand magic tricks, we developed an experimental paradigm to investigate whether magicians’ misdirection techniques could be used to induce the misperception of “phantom” objects. While previous experiments investigating sleight-of-hand magic tricks have focused on creating false assumptions about the movement of an object in a scene, our experiment investigated creating false assumptions about the presence of an object in a scene. Participants watched a sequence of silent videos depicting a magician performing with a single object. Following each video, participants were asked to write a description of the events in the video. In the final video, participants watched the Phantom Vanish Magic Trick, a novel magic trick developed for this experiment, in which the magician pantomimed the actions of presenting an object and then making it magically disappear. No object was presented during the final video. The silent videos precluded the use of false verbal suggestions, and participants were not asked leading questions about the objects. Nevertheless, 32% of participants reported having visual impressions of non-existent objects. These findings support an inferential model of perception, wherein top-down expectations can be manipulated by the magician to generate vivid illusory experiences, even in the absence of corresponding bottom-up information.
Article
Forcing is usually described as the effect in which stage magicians covertly influence decisions made by spectators. The phenomenon has been subject to a number of recent articles and is typically placed within the context of social influence, priming, decision making, awareness, free will, and the science of magic. In the present paper I will argue that forcing researchers, when framing and describing the phenomenon, have exaggerated what magicians typically achieve with the technique. Specifically, the magician is said to influence and manipulate the spectator’s decision when in fact the vast majority of forces do not include any such influence. The consequence of this misrepresentation is that psychologists will be led to believe that the forcing phenomenon has more to contribute to priming and the psychology of influence than it actually does.
Article
Full-text available
Drawing inspiration from sleight-of-hand magic tricks, we developed an experimental paradigm to investigate whether magicians’ misdirection techniques could be used to induce the misperception of ‘phantom’ objects. While previous experiments investigating sleight-of-hand magic tricks have focused on creating false assumptions about the movement of an object in a scene, our experiment investigated creating false assumptions about the presence of an object in a scene. Participants watched a sequence of silent videos featuring a magician performing with a single object. Following each video, participants were asked to write a description of the events in the video. In the final video, participants watched the Phantom Vanish Magic Trick, a novel magic trick developed for this experiment, in which the magician pantomimed the actions of presenting an object and then making it magically disappear. No object was presented during the final video. The silent videos precluded the use of false verbal suggestions, and participants were not asked leading questions about the objects. Nevertheless, 32% of participants reported having visual impressions of non-existent objects. These findings support an inferential model of perception, wherein top-down expectations can be manipulated by the magician to generate vivid illusory experiences, even in the absence of corresponding bottom-up information.
Article
Full-text available
This study examined parental encouragement of imaginative involvement, pathological dissociation, and nonpathological dissociative tendencies (psychological absorption) as predictors of paranormal belief. In a questionnaire survey, data on each of these variables were solicited from 139 Australian university students. Canonical correlation analyses indicated that parental encouragement of imagination was related to relatively "occult" paranormal beliefs such as witchcraft; this association was not mediated by dissociative tendencies. Additionally, pathological dissociation was found to predict belief in parapsychological and spiritual concepts, but psychological absorption was not a correlate of any paranormal belief. Future models of paranormal belief need to envisage different developmental pathways for distinct facets of paranormal belief.
Chapter
It is clear that a wide range of situations exist that can potentially lead people to believe that they have experienced the paranormal when in fact they have not. The question regarding possible differences between believers and non-believers in the paranormal in terms of proneness to cognitive biases can now be answered rather more definitively than has been possible previously. Believers in the paranormal tend to be poorer at syllogistic reasoning, have a more distorted concept of randomness leading them to see meaning where there is none, are more susceptible to experiencing anomalous sensations and are, in certain circumstances, more suggestible. Memory biases in the accuracy of eyewitness testimony for ostensibly paranormal events have also often been reported, and evidence is beginning to accumulate that believers may be more prone to false memories.
Article
The right of Christopher C. French to be identified as the owner of this work has been asserted by him in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved Copyright throughout the world No part of this publication may be reproduced or transmitted in any form or by any means, electronic, mechanical or photographic, by recording or any information storage or retrieval system or method now known or to be invented or adapted without prior permission obtained in writing from the publisher, The Institute for Cultural Research, except by a reviewer quoting brief passages in a review written for inclusion in a journal, magazine, newspaper or broadcast.