A validation study of the Harvey Impostor Phenomenon Scale.

Abstract

This article examines the construct validity of the impostor phenomenon, as measured by the Harvey Impostor Phenomenon Scale (HIPS). A factor analysis was performed on the fourteen scale items, and the results revealed that three factors accounted for 54.7% of the variance. The six items loading into the largest of these factors identify it as an "impostor" factor, whereas the remaining two factors corresponded to feelings of unworthiness and inadequacy. The results point to the potential usefulness of the factoranalyzed HIPS as a measure of the impostor phenomenon and for the general construct validity of the phenomenon.Several psychoanalysts have used the term "impostor" to describe a male who attempts to reconcile the inadequate person he is with the masculine ideals conveyed to him by assuming false identities and by fabricating his achievements. This impostor suffers from a profound impairment of his sense of identity, knows he is not the person he pretends to be, and assumes a false identity for the sole purpose of deceiving others (Chasseguet-Smirgel, 1985;Kaplan, 1984).

A Psychometric Evaluation of the Harvey Imposter Phenomenon Scale

“…Furthermore, Edwards et al reported an alpha reliability for the total score of .34 yet argued for acceptable alpha levels for each of the three retained components (.81, .71, and .65, respectively). Ultimately, Edwards et al (1987) concluded that their study provides strong evidence for the validity of the Harvey IP Scale; yet this conclusion is not supported based on their own criteria of simple structure.…”

“…Gorsuch, 1983). More specifically, Edwards et al (1987) reported dual coefficient loadings on two items (Items 1 and 14, respectively) and no salient loadings for Item 12. Furthermore, Edwards et al reported an alpha reliability for the total score of .34 yet argued for acceptable alpha levels for each of the three retained components (.81, .71, and .65, respectively).…”

“…The Kaiser criterion rule (K1) in which eigenvalues greater than 1 and Cattell's (1966) scree test were initially examined to estimate the number of retained components, as these criteria have typically been reported in previous studies (cf. Edwards et al, 1987;Fried-Buchalter, 1992). It is important to note that Zwick and Velicer (1986) demonstrated the K1 criterion overestimated the number of sample components relative to a generated population structure.…”

“…Relative to internal consistency reliability, the Harvey IP Scale has been varied with coefficient alpha scores in acceptable ranges between .73 and .91 (e.g., Fried-Buchalter, 1992;Holmes, Kertay, Adamson, & Clance, 1993;Ross & Krukowski, 2003;Topping, 1983;Topping & Kimmel, 1985) as well as poor reliability scores of .34 (Edwards, Zeichner, Lawler, & Kowalski, 1987) and .64 (Kolligian & Sternberg, 1991). Although the reliability of scores obtained using the Harvey IP Scale is greatly varied, minimal consideration has been applied to the factor structure of scores using this instrument.…”

“…In an attempt to estimate the IP's factor structure, Edwards et al (1987) conducted a factor analysis on the Harvey IP Scale using 104 postgraduate students. Edwards et al determined that three factors could be extracted with eigenvalues greater than 1 (i.e., K1 rule) to explain the structure of correlations among the 14-item scale.…”