No abstract

Research with Rotter's I-E scale has generally proceeded on the assumption that the scale measures a unidimensional trait. The present study explored the tenability of this assumption by factor analyzing the 23 I-E item responses given by 159 college males and 157 college females. Two factors were identified: a belief concerning felt mastery over the course of one's life, and a belief concerning the extent to which the individual citizen is deemed capable of exerting an impact on political institutions. It was suggested that predictions involving the I-E scale might be refined by separate consideration of the two factors.In his well-known monograph, Rotter (1966) introduced the Internal-External (I-E) scale as a measure of individual differences in the generalized belief that a person can control his own destiny. Considerable research on diverse populations has demonstrated the construct validity of the scale in a variety of experimental and field situations (Lefcourt, 1966;Rotter, 1966). However, correlations between I-E scores and other variables have often been disappointingly low. Rotter has speculated that the failure to obtain stronger relationships may be attributable, at least in part, to the influence of situation specific attitudes in the scale items and in the areas of behavior investigated (Rotter, 1966, p. 21).Two factor analyses of the I-E scale are mentioned in the monograph. One of these, carried out by Rotter on data from 200 male and 200 female college students, indicated the presence of a general factor accounting for "much of the variance" in item responses. Several additional factors, reported as accounting for "small but significant variance," were highly specific, involved only two or three items each, and did not appear to warrant the identification of clear-cut subscales. 2 The second factor analysis of the I-E scale, originally reported by Franklin (1963), employed a principal components analysis on data obtained from 1,000 high school students. This analysis yielded an unrotated factor ac-1 Requests for reprints should be sent to Herbert L. Mirels, Department of Psychology, 1945 High Street, Ohio State University, Columbus, Ohio 43210. 2 The method used to extract factors was not reported.counting for 53% of the total scale variance. Loadings of the individual test items on this factor were uniformly low, with only six of them equal to or exceeding ±.30. The first seven factors obtained by the unrotated solution were rotated to simple structure by the Carroll Biquartimin Oblique Solution (Carroll, 1957). Although the rotated factors appeared to suggest the presence of I-E subareas, none of the factors contained more than one item loading higher than ±.31. In fact, 42 of the 48 item loadings characterized by Franklin as important to factor description failed to exceed ±.20.The present study was carried out in an attempt to clarify the factor structure of the I-E scale, with the anticipation that subsets of items, should they be found to cluster meaningfully, might be employed separate...

It was suggested that fluency, defined as number of responses, may misleadingly influence both high Intercorrelations sometimes reported among measures of creativity and low correlations sometimes reported between measures of creativity and intelligence. Subjects were 93 Saturday art school students between the ages of 9 and 15 years. Intercorrelations among five “creativity” scores derived from a slightly modified version of Torrance's Figure Completion Test and between these scores and Henmon‐Nelson Intelligence were compared using both raw creativity scores and creativity scores corrected for the effect of fluency. Uncorrected creativity scores intercorrelated high among themselves (mean r= .45) and low with intelligence (mean r= .09), while corrected creativity scores showed low intercorrelations among themselves (mean r= .08) and with intelligence (mean r= .13). These findings are interpreted as confirming the influence of fluency upon high intercorrelations among so‐called measures of creativity and as failing to support the suggestion that fluency may also influence low correlations among creativity and intelligence measures.