A Process Dissociation Approach to Objective-Projective Test Score Interrelationships

Abstract

Even when self-report and projective measures of a given trait or motive both predict theoretically related features of behavior, scores on the 2 tests correlate modestly with each other. This article describes a process dissociation framework for personality assessment, derived from research on implicit memory and learning, which can resolve these ostensibly conflicting results. Research on interpersonal dependency is used to illustrate 3 key steps in the process dissociation approach: (a) converging behavioral predictions, (b) modest test score intercorrelations, and (c) delineation of variables that differentially affect self-report and projective test scores. Implications of the process dissociation framework for personality assessment and test development are discussed.

Evidence of Incremental Validity between Pfister Test and Human Figure Drawing

“…Thus, the lo w correlation of scores for d ifferent measures for evaluating the same construct does not necessarily mean a problem of incremental valid ity. Th is is because somet imes the lack of weak co rrelat ion o r lo w magnitude is evidence to support the valid ity of pro ject ive techniques, demonstrating its clinical and empirical usefulness for certain constructs [11].…”

Aggressive Personality: The Use of a New Self-Report Measure Built from Justification Mechanisms

“…This finding was quite surprising considering the advantages stated in the literature about using implicit and explicit cognitions to capture personality and predict behaviors Bornstein, 2002;Winter et al, 1998). However, the literature also states that due to low power for detecting interactions, researchers should consider increasing power by setting critical alpha at .10 for the interaction term in MHMR analyses (Cohen, 1988;Champoux & Peters, 1980;Finn & Frone, 2004;McClelland & Judd, 1993).…”

“…There is significant difference between implicit and explicit cognitions (Bornstein, 2002;Brewin, 1989;Epstein, 1994;Fazio & Olson, 2003;Greenwald & Banaji, 1995;Hogan, 1991;James, 1998;LeBreton, Binning, & Adorno, 2006;McClelland et al, 1989;Mischel & Shoda, 1995;Spangler, 1992;Westen, 1998;Winter et al, 1998). Specifically, as McClelland and colleagues (McClelland, 1985;McClelland et al, 1989) found, when implicit and explicit personality cognitions are measured they are often uncorrelated and are likely to interact when predicting various criteria.…”

“…Research shows the added benefit of using an explicit measure and an implicit measure in combination to fully capture personality constructs (Bing, Stewart, Davison, Green, McIntyre, & James, 2007;Bing, LeBreton, Davison, Migetz, & James, 2007;Bornstein, 2002;James, 1998;Winter, John, Stewart, Klohnen, & Duncan, 1998). Selfreport measures of conscious (explicit) cognitions have a history of low and often nonsignificant correlations with various implicit, usually projective, measures (Greenwald & Banaji, 1995;McClelland, Koestner, & Weinberger, 1989;Nisbett & Wilson, 1977).…”

“…(Greenwald & Banaji, 1995;McClelland, Koestner, & Weinberger, 1989;Nisbett & Wilson, 1977). New theories regarding the interaction between explicit and implicit measurements have been investigated by Bornstein (2002;process dissociation models) and Winter, John, Stewart, Klohnen, and Duncan (1998;channeling models). Evidently, there is great advantage to using both measurement methods when assessing personality.…”