You are on page 1of 4

Psychological Reports, 1987, 61, 435-438.

@ Psycholog~calReports 1987

DO THE BARNUM EFFECT A N D P A R A N O W BELIEF


INVOLVE A GENERAL GULLIBILITY FACTOR?'

LIONEL STANDING AND GREGORY KEAYS


Bishop's Univmity

Summary.-A questionnaire survey indicated that paranormal belief


among 60 college students is positively correlated with over-all levels of be-
lief in a variety of everyday statements ( r = .56), whereas the Barnum effect
shows no such indication of a general gullibility factor. Also, the Barnum
effect is not associated with paranormal belief.

The habitual tendency of individuals to agree that spurious personality


profiles describe them accurately, known as the Barnum effect, was first noted
by Forer (1949). Clinical patients who believe erroneously that their per-
sonalities are exceptionally wicked or disturbed may be displaying a parallel
process of failure to discriminate between unique and universal attributes.
Despite many replications, the Barnum effect remains largely unexplained
(Dickson & Kelly, 1985).
Another striking instance of widespread and unfounded belief is found
in the area of anomaliscic psychology. Many surveys show that high levels of
paranormal belief, involving myriad alleged occult and supernatural events,
occur at even the higher educational levels (Morris, 1981).
Both phenomena must involve a tendency towards what may be termed
gullibility. Is this a general trait, and are the two phenomena related to each
other? The present study tests whether levels of belief in these mro areas are
associated with each other, as well as w ~ t hacceptance of statements in other
areas of general knowledge. The occurrence of such correlations would indi-
cate the possible operation of a general gullibility factor.
METHOD
Szlbjects
Sixty undergraduates registered in psychology courses were employed as
volunteer subjects.
1Materials
An 88-item questionnaire was devised to measure levels of belief, which
presented eight statements in each of 11 subject areas (paranormal phenomena,
geography, religion, science, local events, news stories, psychology, politics,

'This smdy was supported financially by the Bishop's University Research Cornmirree.
The help of S. Black, A. de Man, and S. McKelvie is gratefully acknowledged. Re-
prints may be obtained from L. Standing, Department of Psychology, Bishop's Uni-
versity, Lennoxville, PQ, Canada JIM 127.
436 L. STANDING & G. KEAYS

history, scientific method, and general trivia). The questionnaire was pre-
sented as a cest of general knowledge, but the statements were chosen to be
generally too difficult for the average student (e.g., "Kingston is on the
Rideau river", "Tiberius Caesar reigned during the life of Christ", "Sounds
in water go slower than in air", "The war of 1812 lasted only six weeks").
Consistent posicive responding may be taken to indicate positive response
bias or gullibility, since half of the items were objectively true, as defined by
reference sources, while half were false. Order was randomized. The subject
responded to each scatemenc on a 7-point scale ( 1 false, 7 true).
A Barnilm profile ( a personality sketch caken from Forer, 1949, p. 119)
nlas also employed, plus a list of 20 random personality traits (e.g., talkative,
kind, sarcastic, outgoing).
Procedure
The subjects were first given the list of 20 traits and checked those traits
which chey felt they possessed; they were cold that a computer would analyze
their responses to discover their individual personalities. Subjects also indi-
cated cheir rating of acceptance for each statement in the belief quescionnaire
at this session.
Two days later, all subjects were given an identical personality descrip-
tion (the Barnum profile), which they were told had been generaced by com-
puter analysis of their self-reported traits. They then rated it on che three
7-point scales previously employed by Scanding and Keays ( 1986). These
scales asked: how accurate is che profile, how much insight does it provide,
how well does it describe the may people see you? ( 1 totally inaccurate, 7
totally accurate).
Anonymity was provided for the subjects through the use of pseudonyms.

The three individual scales of Barnum belief were all significantly inter-
correlated, with Pearson coefficients of approximately .60 ( p < .01) and
were therefore summed to provide a composite index of che Barnum effect for
each subject ( M = 4.77, SD = 1.08). Scores within the 11 sections of the
belief questionnaire were also summed to provide an index of belief in each
area of knowledge.
The Barnum index was totally uncorrelated with belief in paranormal
events ( r = -.001) and with total belief scores obtained by summing all 11
sections of the belief questionnaire ( r = .07). It also showed no association
wich the scores for each of che individual sections, the highest correlation
being .19; the mean r was .02.
Paranormal belief was unrelated to scores on the other ten sections of
the belief questionnaire when these were examined individually, the mean
BARNUM EFFECT AND PARANORMAL BELIEF 437

correlation being .13. However, paranormal belief was significantly associated


with a higher total belief score (the sum of the remaining sections of the
questionnaire); r = .56 ( p <.01). This total belief score correlated, on
average, at .44 with the individual section scores, so the data suggest a general
gullibility factor which enters into the paranormal area to about the same
degree as the remaining belief areas.
True versus false items were also examined across the questionnaire; the
pattern of correlations remained unchanged. However, mean scores for the
true items were higher in every subject area (grand means of 4.71 versus 3.10;
SD of .43 and .45), indicating that subjects were not guessing randomly.
The mean scores for paranormal belief were close to the means for all
other areas given above (4.30 and 3.62, for true and false items, SD 1.04 and
1.20, respectively). Credulity appears to be high over-all, rather than being
concentrated in the paranormal area.
Drscussro~
The Barnum effect appears to be specific to the personality-judgement
situation. By contrast, paranormal belief increases among those individuals
who 'believe' in various other topic areas (taken as a group), suggesting the
operation of a general gullibiliry factor. Furthermore, the Barnum effect and
paranormal belief appear to be totally unrelated to each other.
Standing and Keays (1936) have shown that the Barnurn effect still oc-
curs, undiminished, when falsified self-reported trait lists are shown to the
subject, rather than interpretive profiles. In their data, there was a significant
correlation between the subjects' rating of accuracy of their (spurious) pro-
files and of their (spurious) lists of traits ( r 3 0 = .4, $ < .02). For those
tasks that involve self-judgement of personality, a common gullibility factor
may operate.
However, the present study also indicates that the Barnum tendency of
individuals is not related to their credulity in a more general context: our
acceptance of statements about our own personalities is uncorrelated with our
acceptance of statements about the external world. This result appears con-
sonant with the finding by Johnson, at al. (1985) that cognitive bias (judge-
mental perspective) and motivational factors (attractiveness of positive state-
ments) both contribute to the Barnum effect; neither of these factors is likely
to apply to the judgement of general-knowledge statements.
REFERENCES
DICKSON, D. H., & KELLY, I. W. ( 1 9 8 5 ) The 'Baroum effect' i n personality assess-
ment: a review of the literature. Psychological Reports, 5 7 , 367-382.
FORER,B. R. (1949) The fallacy of personal validation: a classroom demonstration of
gullibility. Journal of Abnormal and Social Psychology, 44, 118-123.
JOHNSON, J. T., CAIN,L., FALKE,T.,HAYMAN, J., & PERILLO,E. (1985) The "Bar-
n u m effect" revisited: cognitive and motivational factors in the acceptance of
438 L. STAND~NGa G. KEAYS

personality descriptions. Journal o f Personality and Social Psychology, 49, 1378-


1391.
~.
MORRIS,S. (1981) Believing in ESP: effects of dehoaxing. In K. Frazier (Ed.),
Paranormal borderlands o f science. Buffalo, NY: Prometheus.
STANDING, L., & KEAYS, G. (1986) Computer assessment of personality: a demon-
stration of gullibility. Socal Behavior and Personality, 14, 197-202.

Accefiied August 26,1987.

You might also like