The underlying mechanisms of the Affect Misattribution Procedure
This project investigates the underlying mechanisms of the Affect
Misattribution Procedure (AMP; Payne et al., 2005) with the use of
experimental designs able to create dissociations between affective and
semantic processes. The results imply that the AMP could be used to
measure constructs other than implicit attitudes.
Unlike other variants of affective priming paradigms (e.g., Fazio,
Sanbonmatsu, Powell, & Kardes, 1986; Murphy & Zajonc, 1993) the
AMP not only produces strong and robust effects, but also has good
psychometric properties (internal consistency: .69 < ? < .90;
Payne et al., 1995; Payne, Burkley, & Stokes, 2008). Among the many
latency-based indirect measures, only the Implicit Association Test
(IAT, Greenwald, McGhee, & Schwartz, 1998) reaches this level of
reliability. Consequently, the AMP quickly became popular among social
cognition researchers who were interested in using a second,
conceptually different, measure of implicit attitudes. What is less
clear are the underlying processes of the AMP. To test three different
theoretical accounts for the AMP effect: affect priming, affect
misattribution, and semantic priming, I created a modified version of
the Affect Misattribution Procedure (AMP; Payne et al., 2005). Three
studies showed consistent evidence that the AMP effect was mediated by
semantic priming (Blaison, Imhoff, Hess, & Banse, 2010). At
variance with the theoretical explanation by Payne and colleagues,
evidence for an affect-driven process was less consistent as only a
weak effect of affect priming, and no significant effect of affect
misattribution emerged. Overall, the results are compatible with a
semantic account of AMP effects. This interpretation implies that the
scope of application of the AMP in implicit social cognition research
may be much broader than previously assumed.