Beruflich Dokumente
Kultur Dokumente
To cite this article: Lynn E. McCutcheon , Diane D. Ashe , James Houran & John Maltby (2003) A
Cognitive Profile of Individuals Who Tend to Worship Celebrities, The Journal of Psychology, 137:4,
309-322, DOI: 10.1080/00223980309600616
309
31 0 The Journal of Psychology
I don’t care who he dates, sleeps with, marries . . . I realized I wanted him to be
happy, and that that would make me happy . . . He’s the only person I connect with
(Haynes & Rich, 2002, p. 198).
Method
Participants
Measures
algebra. An example is “If Tom had 4 times as much money as John, he would
have $24. How much money does John have?’ The items on both subscales
become progressively more difficult. One point was scored for each correct
answer; thus Information scores could range from 0 to 40, and Arithmetic scores
could range from 0 to 26. We chose those two subscales because they could be
administered in a short period of time and because they were designed to tap dif-
ferent kinds of cognitive skills.
We selected 20 of the 43 items from Version 2 of the Scottsdale Test of
Critical Thinking (Ricker, 2001). We selected a critical thinking measure
because these kinds of measures are designed to tap a respondent’s ability to
use fluid intelligence. Preliminary work has shown that the Scottsdale Test of
Critical Thinking has adequate reliability and validity (Ricker, personal com-
munication, September 9, 2001). Each item presents a scenario followed by
four choices. Items are designed to illustrate common errors in critical thinking,
such as the ad hominem fallacy, appeal to authority, generalization from small
sample size, and circular reasoning. A representative item is “John now tells
people to take X because a doctor once told him that X prevents asthma. Which
of the following criticisms is the BEST one to make regarding John’s reason-
ing?’ The correct answer is that his decision is based on the doctor’s authority.
We chose only 20 of the items for inclusion because of their length and diffi-
culty; each correct answer was worth one point.
The Squares Problem was selected because it is a nonverbal measure that taps
spatial ability, and it is easily administered. We do not know the origin of the prob-
lem, but we believe it has appeared in more than one book of puzzles. A large
square is presented; this square consists entirely of 16 small squares. Two smaller
squares are superimposed within so that their centers are to the immediate left and
immediate right of the intersection that lies at the exact center of the largest square.
The two smaller squares are equal in size and comprise four tiny squares each.
The respondent’s task is to correctly identify the total number of squares-
in this example, the answer would be 40 squares. For the purpose of the present
study, the score was the number recorded by the participant at the conclusion of
a 5-min interval. Some participants apparently solved part of the problem and
then gave an estimated answer. Some of these answers were higher than 40. We
adjusted these (so that our scoring system made sense) by evaluating each incor-
rect answer in relation to the correct one. Thus, an answer of 43 became 37, 42
became 38, and 41 became 39.
The Need for Cognition Scale (Cacioppo & Petty, 1982) is different from
our other measures because it assesses attitude rather than ability. We used the
18-item version (Cacioppo, Petty, & Kao, 1984) with responses ranging from
extremely uncharacteristic of me (1) to extremely characteristic of me (5). Half
the items, such as “Thinking is not my idea of fun,” are reverse scored. High
scorers tend to relish the opportunity to solve complicated puzzles and enjoy the
challenge posed by intellectual problems. The Need for Cognition Scale has
McCutcheon, Ashe, Houran, & Maltby 3 I5
good reliability and validity and has been found to correlate moderately with
general intelligence (Cacioppo & Petty, 1982; Cacioppo et al., 1984).
Procedure
The Squares Problem and the Need for Cognition Scale were completed as
part of regular classroom activities in critical thinking classes taught by the first
author. We allotted exactly 5 min for the former and approximately 5 min for the
latter. Because of classroom absences, sample sizes for the Squares Problem and
the Need for Cognition Scale were 82 and 90, respectively.
All the participants were recruited from these classes with the inducement of
extra credit. They reported to a vacant classroom in eight groups ranging in size
from 7 to 20. They were given a booklet containing all the measures except for
the Squares Problem and the Need for Cognition Scale and were asked to com-
plete the Celebrity Attitude Scale and accompanying demographic questions.
When the participants had finished this part under no time restrictions, the
door was locked to avoid distractions. The instructions for each of the measures
were read prior to testing. We told the participants that they would have 7 min for
the Arithmetic and Information subscales and 15 min each for the Remote Asso-
ciates Test and the Scottsdale Test of Critical Thinking. They were prompted
when 10 s remained in each test period so that they could record a last answer.
We used eight different orders of presentation of the four ability measures to min-
imize the probability of a systematic order effect. We asked the participants not
to discuss any testing materials with anyone. They were debriefed a few weeks
later in class.
Results
Preliminaries
Measure a M SD Range
"The original score sheets for our sample were accidentally discarded. Therefore, this coeffi-
cient represents the split-half reliability reported by the authors of the Need for Cognition
Scale (Cacioppo & Petty, 1982).
greater relevance than others, but a correlational analysis suggested that this find-
ing does not apply to our measures of cognitive abilities.
We were struck by the remarkable consistency revealed in Table 2. All six
cognitively related variables were correlated negatively with total scores on the
Celebrity Attitude Scale-and four were highly significant (p < .oO1 ) and
remained robust after we applied a Bonferroni correction for multiple analyses. In
other words, our results showed that individuals who were more strongly attract-
ed to their favorite celebrity (celebrity worshippers) were likely to score lower on
measures of general information, creativity, critical thinking, and spatial ability.
Furthermore, all of the cognitive measures were intercorrelated as expected, lend-
ing further support to the claims of validity made for each of these instruments.
Scottsdale
Remote Test of Need for
Associates Critical Squares Cognition
Measure Test Information Arithmetic Thinking Problem Scale
~ ~ ~ ~~ ~ ~
Celebrity Attitude Scale-total score -.39** -.33** -.18 -.42** -.3 1 * -.2 I
Remote Associates Test - .60** .46** .52** .37** .18
Information - .58** .54** .27 .28*
Arithmetic - .47** .32* .29*
Scottsdale Test of Critical Thinking - .I6 .37**
Squares Problem - .06
* p < .01. **p < ,001, significant after Bonferroni adjustment for multiple analyses
3 18 The Journal of Psychology
The multiple correlation R for the regression was significantly different from
zero, R = .56, R2 = .31, F(6, 69) = 5.11, p < .001; that is, total scores on the
Celebrity Attitude Scale were significantly related to the set of independent vari-
ables. Altogether, scores on the six cognitive measures predicted 31% (or 25%
adjusted, adjusted R2 = .25) of the variability in the total scores on the Celebrity
Attitude Scale. In addition, given that the six cognitive measures were intercor-
related, it is important to inspect so-called tolerance statistics (Tabachnick &
Fidell, 1996) and ensure that intercorrelations among predictor variables do not
compromise the analyses. All tolerance statistics in the regression were well
above zero, ranging from .52 to $7; therefore multicollinearity of the predictor
variables was of no practical concern (Darlington, 1990).
Only the Remote Associates Test showed a suggestive trend to contribute
significantly to the prediction of scores on the Celebrity Attitude Scale, although
none of the six cognitive measures individually reached statistical significance at
the .05 level: the Remote Associates Test (creativity) (p = -.24, p = .078), Infor-
mation (p = -.17,p = .225),Arithmetic (p = .09, p = .445), the Scottsdale Test of
Critical Thinking (p = -.22, p = .116), the Squares Problem (spatial ability) (p
= -.15, p = .164), and the Need for Cognition Scale (p = -.04,p = .700).
We performed three additional standard multiple regression analyses to see
whether these statistical trends could be replicated for each of the three subscales
of the Celebrity Attitude Scale.
the variance in the scores. We observed only a suggestive trend for scores on
the Scottsdale Test of Critical Thinking to contribute significantly to the pre-
diction of scores on this subscale, but again none of the six cognitive measures
individually reached statistical significance at the .05 level: the Remote Asso-
ciates Test (creativity) (p = -.19, p = .173), Information (p = -.21, p = .149),
Arithmetic (p = .01, p = .915), the Scottsdale Test of Critical Thinking (0 =
-.23, p = .092), the Squares Problem (spatial ability) (p = -.04, p = .689), and
the Need for Cognition Scale (p = -.03, p = .768). All tolerance statistics in the
regression were well above zero, ranging from .52 to .87, so multicollinearity
of the predictor variables was of no practical concern.
Discussion
Our findings might indicate simply that people with higher cognitive function-
ing are protected from becoming absorbed in fantasies about celebrities for various
reasons. These individuals may understand that the celebrity system is in essence
designed to sell a product; or these individuals may be more interested in ideas than
in people and their personalities. They may also feel they are as talented or bright as
the celebrities being promoted by the media and entertainment industry.
However, the resemblance between an erotomanic type of delusional disor-
der and the phenomenon of celebrity worship is bolstered by the present findings,
which conceptually replicate neuropsychological profiles of patients diagnosed
320 The Journal of Psychology
with erotomania (cf. Fujii et al., 1999). In particular, high scores on the Celebri-
ty Attitude Scale consistently showed negative correlations with our measures of
cognitive ability. Standard multiple regression analyses revealed that only the
cumulative effect of the cognitive measures predicted total scores on the Celebri-
ty Attitude Scale, and this trend was replicated in regression analyses that con-
sidered individually the three subscales of the Celebrity Attitude Scale (Enter-
tainment-Social, Intense-Personal, and Borderline Pathological).
The only exception to this trend was the small but significant contribution
of scores on the Remote Associates Test (creativity) in predicting scores on the
Borderline Pathological subscale. These findings implicate a broad range of cog-
nitive deficits in celebrity worship and specifically identify in borderline patho-
logical celebrity worshippers a deficit in the ability to form verbal remote asso-
ciations. This latter finding is consistent with the Fujii et al. (1999) report that
erotomanic patients exhibit deficits in working memory, cognitive flexibility,
and the ability to form new verbal associations.
The present results extend the absorption-addiction model of celebrity wor-
ship (McCutcheon et al., 2002) by elucidating several cognitive factors that coin-
cide with an individual’s tendency to become engrossed with the fantasy of
celebrity. However, the modest effect sizes in this study indicate that cognitive
deficits alone are insufficient to account for attitudes and behaviors that define
celebrity worship. Rather, such deficits may simply facilitate an individual’s sus-
ceptibility to engage in this behavior.
For example, we speculate that visuospatial-functioning deficits, in combi-
nation with isolative and ambivalent interpersonal relationships, also partly stem-
ming from cognitive deficits, may contribute to misinterpretations in parasocial
relationships. Deficits in cognitive flexibility-in addition to the proposed role of
an addictive element-may subsequently contribute to the maintenance of the
delusional belief. Fuji et al. (1999) proposed similar mechanisms to account for
erotomania. Therefore, it is important for future researchers to determine whether
the two phenomena are distinct clinical entities or merely different expressions
of a common neuropsychological state.
Additional study of these issues might reveal a set of diagnostic criteria for the
celebrity worshipper that can help guide the development of effective therapeutic
strategies for use in clinical contexts. It is unrealistic to expect that the field can fully
combat the phenomenon of celebrity worship, but we hope to understand the devel-
opment and progression of this phenomenon well enough to address extreme
episodes such as the kind exhibited by the female adolescent who was willing to
mutilate herself over her parasocial relationship with a popular musician.
REFERENCES
Aiken, L. R. (1997). Psychological resting and assessment (3rd ed.). Needham Heights,
MA: Allyn & Bacon.
McCutcheon, Ashe, Houran, & Maltby 321
McCutcheon, L. E. (2002). Are parasocial relationship styles reflected in love styles? Cur-
rent Research in Social Psychology, 7, 82-94 [online]. Retrieved April l , 2002, from
http://www.uiowa.edu/-grpproc/crisp/crisp.7.6.htm
McCutcheon, L. E., Lange, R., & Houran, J. (2002). Conceptualization and measurement
of celebrity worship. British Journal of Psychology, 93, 67-87.
McCutcheon, L. E., & Maltby, J. (2002). Personality attributions about individuals high
and low in the tendency to worship celebrities. Current Research in Social Psychology,
7, 325-338 [online]. Retrieved April 1, 2002, from http://www.uiowa.edu/-grpp
roc/crisp/crisp.7.19.html
Mednick, S. A., & Mednick, M. T. (1967). Examiner’s manual: Remote Associates Test.
Boston: Houghton Mifflin.
Raviv, A., Bar-Tal, D., Raviv, A., & Ben-Horin, A. (1996). Adolescent idolization of pop
singers: Causes, expressions, and reliance. Journal of Youth and Adolescence, 25,
63 1-650.
Ricker, J. P. (2001). Scottsdale Test of Critical Thinking-Version 2. Unpublished manu-
script available from the author at Scottsdale Community College, Scottsdale AZ. E-
mail: Jeff.Ricker@SCCmail.maricopa.edu.
Showalter, E. ( 1 997). Hystories: Hysterical epidemics and modem media. New York:
Columbia University Press.
Stever, G. S. (1991). The Celebrity Appeal Questionnaire. Psychological Reports, 68,
859-866.
Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics (3rd ed.). New
York: HarperCollins.
Yue, X. D., & Cheung, C.-K. (2000). Selection of favourite idols and models among Chi-
nese young people: A comparative study in Hong Kong and Nanjing. International
Journal of Behavioral Development, 24, 9 1-98.