Академический Документы
Профессиональный Документы
Культура Документы
SUKHDEV SINGH
Directorate of Education
Delhi Administration, Delhi
AMPLE EVIDENCE (e.g.. Miller & Chapman, 1983; Upmanyu, Gill, & Singh,
1982; Upmanyu & Upmanyu, 1988) has indicated that schizophrenics' respons-
es on the Word Association Test (K-R WAT; Kent & Rosanoff, 1910) are more
unusual than those of normal participants. However, one problem in equating
unusual word associations with psychiatric malfunction is that unusual word
associations can also be indicative of creativity (Barron, 1965, 1969; Guilford,
Wilson, Christensen, & Lewis, 1951; Maltzman, 1960; Mednick, 1962;
The response entropy values for the 100 stimulus words from the Kent-Rosanoff Word
Association Test are available from thefirstauthor.
Address correspondence to V. V. Upmanyu, Department of Psychology, Panjab Uni-
versity, Chandigarh-160014, India.
521
522 The Journal of Social Psychology
Upmanyu, Gill, & Singh, 1982; Upmanyu & Singh, 1984). Gough's (1976) study
indicated that moderately infrequent word associations were more strongly relat-
ed to creativity than extremely remote word associations or common word asso-
ciations were. "From a theoretical perspective, the finding that moderately
unusual replies are critical is worthy of attention. Associations that are less atyp-
ical seem to be indicative of the kind of new perspectives and original reactions
that are almost by definition a part of the creative process" (Gough, p. 352).
However, because psychiatric disturbance was not assessed in Gough's study, it
would be dangerous to link only creativity and not psychiatric disturbance with
even moderately infrequent word associations. The inclusion of creativity mea-
sures and the total exclusion of emotional disturbance measures make discerning
simple creativity correlates of word associations more difficult.
Furthermore, because research findings (Brown, 1965, 1970; Hundal &
Upmanyu, 1981; Innes, 1972; Kuntz, 1974; Laffal, 1955; Paivio, Yuille, & Madi-
gan, 1968; Penk, 1978; Shiomi, 1979; Upmanyu, 1981) have indicated that the
characteristics of stimulus words have a demonstrable effect on word associa-
tions, it would be unwise to assume emotional disturbance or creativity, or both, on
the basis of a single idiosyncratic, unusual, or common word association. However,
although this conclusion may be disquieting to researchers who use verbal materi-
al to study topics such as perceptual defense and repression, it is not uncondition-
al. Studies on personality and cognitive correlates of word associations that fail to
consider the possible infiuence of stimulus-word characteristics may need to be
reevaluated in terms of stimulus variables. Hundal and Upmanyu (1981) and
Upmanyu (1981) suggested that researchers who use word association tests for
diagnostic purposes should control or manipulate the stimulus characteristics.
Thus, the response entropy (a measure of randomness of association) of the
stimulus word, as set forth by Laffal (1955), is an important intervening variable
and must be considered in any proper assessment of the genuineness of word
association emotional indicators as indices of emotionality. Response entropy
encompasses not only the number of different responses to a stimulus word but
also the percentage of participants who respond a certain way.
Method
The participants were 250 male graduate students living in university hostels
in India. The participants were required to be (a) free of any indication of acute
confusion, brain damage, alcoholism, or drug addiction and (b) not currently in
treatment for a diagnosed psychiatric illness. The participants were contacted
individually and were assured of anonymity and confidentiality.
sonality Questionnaire (Eysenck & Eysenck, 1975); the Torrance Test of Creative
Thinking, Figural and Verbal Form A (Torrance, 1966); and the MMPI-Psycho-
pathic Deviate subscale (Hathaway & McKinley, 1967). The psychometric char-
acteristics of these tests have been documented among Indians (Upmanyu &
Singh, 1984; Upmanyu & Upmanyu, 1988).
The tests were administered in random order during five sessions. With the
exception of the Word Association Test, which was administered individually, the
tests were administered to groups of 10 to 15 participants. The participants
received the following instructions:
I am going to read a list of words. After each word I read, say the first word tbat
comes into your mind. The word may or may not be related to the word I read. I want
you to answer as quickly as possible because I am going to record the amount of time
it takes you to respond.
I am going to present the same list of words again. After each word tbat I read, try to
re.spond with the same word you responded witb the first time the list was presented.
For example, if 1 said "bazaar" and you responded "shop" the first time the list was
presented, this time you should say "shop" after I read the word "bazaar." You may
take as long as you wish.
The 100 stimulus words were again read 1 at a time, in the same order as
before. Emotional indicators, such as forgetting and misremembering, were
recorded.
Scoring
Two hundred fifty participants' responses to 100 stimulus words were tabu-
lated, and the number of different responses for each stimulus word was count-
ed. The code number for the emotional indicators for each response was entered
in the appropriate cells. We calculated response entropy (H) for each stimulus
word, using the following formula: H = - Xp^log^p, (Laffal, 1955), where p, is
the probability of occurrence of a given response (p, =//«, where/denotes the
frequency of occurrence of a response word, and n is the number of participants).
Probability values for the various responses to a stimulus word were used to com-
pute - I P| logj /?! for all the responses elicited by the stimulus word. Following
M ± 1 SD, we selected 70 stimulus words of average response entropy to identi-
fy variotis categories of word-association emotional indicators.
Analysis of Data
Results
Reliability
The test-retest reliability coefficient (A' = 75) for the IPAT Anxiety Scale
Questionnaire after a period of about 1 month was .88, and the reliability coeffi-
cients ofthe subtests ofthe anxiety scale questionnaire were between .32 and .70.
The test-retest reliability coefficients for the Eysenck Personality Questionnaire
ranged from .70 (P scale) to .88 (E scale) and were comparable to previously
published values (Chapman et al., 1982; Eysenck & Eysenck, 1975).
The test-retest reliability coefficients for various indices of creativity for 30
participants ranged from .65 to .86. The reliability coefficients for the word-
association emotional indicators ranged between .57 and .80 and in view of the
projective nature of the test were considered satisfactory.
Factor Analysis
Factor 1 from the varimax solution (see Table 1) can be interpreted as ver-
bal creativity; it loaded heavily on four indices of verbal creativity, with fairly
substantial loadings on extremely unique word associations and moderately
infrequent word associations.
Factor 2 loaded on different components that form Cattell's second-order
factor of anxiety. The N scale of the EPQ also loaded on this factor, which is
clearly anxiety. None of the word-association emotional indicators loaded sig-
nificantly on this factor.
Factor 3 loaded strongly on all four indices of figural creativity and social
desirability (the L scale of the EPQ). The variance attributable to these measures
is more or less exhausted on this factor, which can be interpreted as figural cre-
ativity. The nature of the factor loadings raises the question of whether divergent
ability is multidimensional on this particular test or whether figural creativity is
a general factor. The present results seem to suggest that these intratask elements
of creative expression are part of a common factor. Furthermore, the significant
negative loading of the L scale of the EPQ confirmed the expected negative rela-
tionship between the L scale and creativity.
Factor 4 loaded principally on a set of word-association emotional indicators
(extremely unique associations, response repefition, misremembering) and the P
scale of the EPQ, with a moderate loading on psychopathic deviation (MMPI-Pd
subscale).
Factor 5 comprised Factor L (suspiciousness; a component of a second-order
factor of anxiety) and long reaction time as well as repetition of stimulus before
526 The Journal of Social Psychology
responding. The total common variance accounted for by these three measures
was exhausted on this factor, which was interpretable as paranoid insecurity.
Factor 6 suggested a positive association between extraversion and common
word associations.
TABLE 1
Varimax Rotated Factor Loadings
Factor
Variable 1 2 3 4 5 6 7 8 h'-
1. RSBR -.08 .08 .12 -.08 .79 .06 .11 .01 .67
2. UR .56 -.10 -.02 .60 .12 -.04 -.03 .25 .76
3. LRT .21 .05 .24 .11 .47 .29 -.23 .02 .47
4. RR -.05 .00 -.05 .59 .19 -.08 -.23 .39 .60
5. W -.10 -.22 .00 .67 .14 .00 .15 .15 .57
6. Fg. -.21 -.04 .20 -.03 .28 -.23 -.63 .08 .62
7. UR' .89 .00 .08 -.12 -.09 .03 .02 .06 .83
8. LCR -.05 .10 .02 .06 .02 .88 -.05 -.27 .87
9. CR -.32 .10 -.16 .03 .10 .57 .07 -.17 .51
10. P scale .03 .23 .19 .67 -.09 -.06 -.17 -.12 .59
11. E scale .02 -.28 -.22 .30 -.11 .44 -.01 -.09 .43
12. N scale .06 .59 -.10 .26 -.06 -.28 .28 .01 .59
13. L scale .17 .22 .60 -.30 .20 -.14 -.10 -.04 .60
14. % (-) -.04 .62 .04 .18 -.16 -.02 .01 -.07 .45
15. C(-) .00 .59 .01 .00 -.01 .05 -.03 -.11 .36
16. L -.12 .18 .07 -.07 .76 -.22 .09 .00 .69
17. O -.01 .75 .05 -.14 .00 .04 .05 .05 .59
18. Q^ .07 .72 -.07 -.06 .01 .00 -.33 .10 .65
19. Anxiety -.01 .95 .02 -.06 -.02 -.04 .17 .02 .94
20. Pd scale -.05 .46 -.27 .48 .09 .14 .07 .18 .58
21. F-orig. .05 .06 -.79 .00 .12 -.03 .02 .09 .65
22. F-elab. .09 -.14 -.59 .00 .05 .20 .44 .07 .62
23. F-tlue. .14 .05 -.84 -.09 -.11 -.06 -.08 .01 .76
24. F-flex. .08 .01 -.87 -.10 -.13 -.12 -.01 -.02 .81
25. V-orig. .90 .01 -.11 .01 .06 -.07 -.01 -.03 .83
26. V-flue. .94 .00 -.14 .01 -.05 -.01 -.01 .02 .91
27. V-nex. .89 .03 -.07 -.11 -.09 -.01 .05 .04 .82
% contribution 14.43 13.34 11.70 8.21 6.47 6.23 3.78 1.68 65.84
Note. RSBR = repetition of the .stimulus before responding; UR = unique response; LRT = long reac-
tion time; RR = response repetition; W = misremembering; Fg. = forgetting; LCR = less common
respon.se; CR = common response; Q^ = defective integration; C = ego weakness; L = .suspicious-
ness; O = guilt; Qj = frustrative tension; F-orig. = figural originality; F-elab. = figural elaboration; F-
flue. = figural fluency; F-flex. = figural flexibility; V-orig. = verbal originality; V-flue. = verbal flu-
ency; V-flex. = verbal flexibility.
Upmanyu, Bhardwai. & Singh 527
Discussion
Our focus in the present research was the relationship between word-associ-
ation emotional indicators and creativity and major personality dimensions, such
as anxiety, psychopathic deviation, and Eysenckian psychoticism, neuroticism,
and extraversion. The results indicated that word-association emotional indica-
tors were related to various psychopathological traits measuring disposition
toward psychosis and paranoid insecurity.
Extremely unique word associations emerged as a complex measure. A por-
tion of the variance in extremely unique word associations was positively asso-
ciated with verbal creativity. Because moderately infrequent associations con-
tributed substantially to this factor, it could be argued that rare/uncommon but
not bizarre and atypical word associations are related to verbal creativity. From
a theoretical perspective, the finding that the aspect of extremely unique word
associations that is associated with moderately unusual replies is critical for
defining verbal creativity is worthy of attention.
Factor 4, which refers to the second component of extremely unique word
association, was positively associated with response repetition (common in
schizophrenics), misremembering, psychoficism, and psychopathic deviation.
The large contributions of response repetition and misremembering to this com-
ponent of extremely unique word associations may have refiected bizarre, isolat-
ed, and more or less completely atypical word associations. With respect to the
substantial contribution of the psychoticism factor of the EPQ and the Psycho-
pathic Deviate subscale of the MMPI, it can be inferred that the P scale of the
EPQ tapped some features of social anhedonia (e.g., aloofness, social withdraw-
al), social deviance, unconventionality, and mild antisocial behavior. This finding
seems to provide support for Claridge's (1981) and Raine's (1986) suggestion of
a link between psychoticism, mild antisocial behavior, and lack of conformity/
unconventionality rather than the more specific clinical entity of psychopathic
behavior. Although the nature of the P scale on the EPQ has been questioned,
these findings may help clarify the nature of Eysenck's psychoticism dimension
(Eysenck & Eysenck, 1975, p. 5).
The present findings cast doubt on the generally accepted association
between unusual word associations and schizophrenia. Unusual word associa-
tions should be distinguished as meaningful or bizarre, with each category
accounting for unique cognitive abilities.
A major finding ofthe present study is that long reaction time and repetition
ofthe stimulus word before responding had a strong suspicious content. It is con-
ceivable that suspiciousness is one of several features of these two word-associ-
ation emotional indicators; suspiciousness might make a person repeat a stimu-
lus before responding, delaying reaction time.
528 The Journal of Social Psychology
REFERENCES
Barron, F. (1965). The psychology of creativity. In T. M. Newcomb (Ed.), New directions
in psychology II (pp. 3-134). New York: Holt, Rinehart and Winston.
Barron, F. (1969). Creative person and creative process. New York: Holt, Rinehart and
Winston.
Brown, W. P. (1965). Emotional indicators in word association. British Journal of Psy-
chology, 56, 401^22.
Brown, W. P. (1970). Individual differences in associating to neutral and emotional words.
Journal of Consulting and Clinical Psychology, 34, 33-36.
Cattell, R. B., & Scbeier, I. H. (1963). Handbook for the IPAT Anxiety Scale Que.stion-
naire. Champaign, IL: Institute for Personality and Ability Testing.
Chapman, L. J., Chapman, J. P., & Miller, E. N. (1982). Reliabilities and intercorrelations
of eight measures of proneness to psychosis. Joumal of Consulting and Clinical Psy-
chology, 50, 187-195.
Claridge, G. (1981). Psychotieism. In R. Lynn (Ed.), Dimensions of personality. Papers in
honour of H. J. Eysenck (pp. 79-109). Oxford: Pergamon Press.
Eysenck, H. J., & Eysenck, S. B. G. (1975). Manual of the Eysenck Personality Ques-
tionnaire (Junior and Adult). London: Hodder and Stoughton.
Gougb, H. G. (1976). Studying creativity by means of word association tests. Journal of
Applied Psychology, 61, 348-353.
Guilford, J. ?., Wilson, R. C, Christensen, P. R., & Lewis, D. J. (1951). A factor-analytic
study of creative thinking. I. Hypotheses and description of tests (Reports from the Psy-
chological Laboratory No. 4). Los Angeles: University of Southerti California.
Hathaway, S. R., & McKinley, J. C. (1967). Manual of the Minne.wta Multiphasic Per-
sonality Inventory. New York: The Psychological Corporation.
Hundal, P. S., & Upmanyu, V. V. (1981). Nature of emotional indicators in Kent-Rosanoff
Word Association Test; An empirical elaboration. Personality Study and Group Behav-
ior, I, 50-61.
Innes, J. M. (1972). The relationship of word-association commonality response set to
cognitive and personality variables. The British Journal of P.'iychology, 63, 421^28.
Kent, G. H., & Rosanoff, A. I. (1910). A study of association in insanity. American Jour-
nal of Insanity, 67, 37-96; 317-390.
Kuntz, D. (1974). Response faults on word association as a function of associative diffi-
culty and of affective connotation of tbe words. Journal of Consulting and Clinical Psy-
chology, 42. 231-235.
Laffal, J. (1955). Response faults in word association as a function of response entropy.
Journal of Abnormal and Social Psychology, 50, 265-270.
Maltzman, I. (1960). On the training of originality. Psychological Review, 67, 229-242.
Mednick, S. A. (1962). The associative basis of tbe creative process. Psychological
Review, 69, 220-232.
Miller, E. N., & Cbapman, L. J. (1983). Continued word association in hypothetical psy-
cbosis-prone college students. Journal of Abnormal Psychology, 92, 468^78.
Paivio, A., Yuille, J. G., & Madigan, S. A. (1968). Concreteness, imagery and meaning-
Upmanyu, Bhardwaj, & Singh 529
fulness values for 925 nouns. Journal of Experimental Psychology, Monograph Sup-
plement, 71 (\, Pt. 2).
Penk, W. E. (1978). Effects of ambiguous and unambiguous stimulus word differences on
popular responses of schizophrenics. Journal of Clinical Psychology, , 838-843.
Raine, A. (1986). Psychopathy, schizoid personality and borderline/schizotypal personal-
ity variables in a psychiatric group. Personality and Individual Differences, 7, 493-502.
Shiomi, K. (1979). Differences in reaction tinies of extroverts and introverts to Rapport's
word association test. Psychological Reports, 45, 75-80.
Torrance, E. P. (1966). Torrance tests of creative thinking: Norms, Technical Manual,
Princeton, NJ: Personnel Press.
Upmanyu, V. V. (1981). Study of Kent-Rosanoff Word Association Test in relation to
Knponf^e entropy. Personality Study and Group Behavior, I, 1-52.
Upmanyu, V. V., Gill, P. S., & Singh, S. (1982). Nature of unusual responses in
Kent-Rosanoff Word Association Test and Torrance test of creativity. Personality Study
and Group Behavior, 2, 44—53.
Upmanyu, V. V., & Singh, S. (1984). Word association test responses: An index of patho-
logical characteristics or creativity. Personality Study and Group Behaviour, 4, 39-49.
Upmanyu, V. V., & Upmanyu, S. (1988). Utility of word association emotional indicators
for predicting pathological characteristics. Personality Study and Group Behavior, 8,
13-22.