Methodological issues in the validation of implicit measures: Comment on De Houwer, Teige-Mocigemba, Spruyt, and Moors (2009).

Abstract

J. De Houwer, S. Teige-Mocigemba, A. Spruyt, and A. Moors's normative analysis of implicit measures provides an excellent clarification of several conceptual ambiguities surrounding the validation and use of implicit measures. The current comment discusses an important, yet unacknowledged, implication of J. De Houwer et al.'s analysis, namely, that investigations addressing the proposed implicitness criterion (i.e., does the relevant psychological attribute influence measurement outcomes in an automatic fashion?) will be susceptible to fundamental misinterpretations if they are conducted independently of the proposed what criterion (i.e., is the measurement outcome causally produced by the psychological attribute the measurement procedure was designed to assess?). As a solution, it is proposed that experimental validation studies should be combined with a correlational approach in order to determine whether a given manipulation influenced measurement scores via variations in the relevant psychological attribute or via secondary sources of systematic variance. (PsycINFO Database Record (c) 2009 APA, all rights reserved).

More Complex than Previously Thought: New Insights into The Optimal Administration of the Initial Preference Task

“…This difference clearly shows that recognizing one's initials during the IPT changes the correlation with explicit self-esteem (as compared to nonrecognizers). It could be interpreted as indicating that for non-recognizers the IPT Administration of the IPT 211 might measure less unwanted error variance and thus more implicit aspects of selfesteem as compared to recognizers (see Gawronski, LeBel, Peters, & Banse, 2009, for a discussion). These findings, however, call for replication and clarification in future studies.…”

On the controllability of evaluative-priming effects: Some limits that are none

“…He argued that successful faking does not necessarily jeopardise the predictive validity of evaluative-priming effects. Instead, evaluativepriming effects may still validly reflect interindividual differences in spontaneous prime evaluations despite shifts in the overall mean (see Gawronski, LeBel, Peters, & Banse, 2009, for a similar argument). Participants may thus be unable to control the influence of the prime's evaluation on their responses although they succeed in shifting their evaluative-priming scores by an additive constant.…”

How to Administer the Initial Preference Task

“…However, further research needs to examine whether these differences also reflect differences in implicit self-esteem because variations in measurement outcome can also be due to changes in error variance (see Mierke & Klauer, 2003). Hence, an increased effect size on the NLE might be caused by reduced error variance and not because of tapping into a different aspect of implicit self-esteem (for a discussion, see De Houwer et al, 2009;Gawronski, LeBel, Peters, & Banse, 2009). As has been already stated above, we did not control whether there were differences between experimental groups in the degree of awareness of the studies purpose or not.…”

“…Although the IPT has been frequently used in research, methodology papers are still scarce (for exceptions, see Albers, Dijksterhuis, & Rotteveel, 2009;LeBel & Gawronski, 2009). Hence, it is not surprising that the IPT's administration does vary a lot across studies.…”