Implicit Attitudes: Quantitative Semantic Misattribution Procedure

Abstract

Certain mental processes are suggested to exist beyond conscious awareness and control. These processes have often been categorized as implicit, in contrast to explicit, processes, which are readily available to conscious report. Researchers have attempted to measure and assess these implicit processes in a different number of ways. Projective measures, for instance, present ambiguous or unstructured stimuli to respondents, with the assumption that their responses will reveal aspects about their attitudes, personality, and etc. Despite longstanding evidence disfavoring most projective measures, their use in clinical and forensic settings has been remarkably robust. Phrased in terms of modern psychological research, projection might be considered an instance of misattribution, that is, mistaking the source for the effect. People, for example, might misinterpret the transient pleasure of a sunny day as lasting life satisfaction. Such source of confusion is a common feature of events in everyday life; correction to this misattribution demands motivation, awareness, and control of the bias responsible for the misattribution. This complexity makes projective measures hard to design. In the Affect Misattribution Procedure (AMP, Payne et al., Journal of Personality and Social Psychology, 89, 277–293, 2005), for example, participants are briefly presented with an ambiguous pictograph (e.g., a Chinese character) preceded by a prime (e.g., photos of Black or White people). Participants are then asked to rate the pleasantness of the pictograph—the assumption in this paradigm is that evaluations of how pleasant the pictograph is perceived to be, are influenced by the individual’s automatic affective reactions that s/he has towards the prime. Despite the fact that the AMP is easily administered, shows good internal consistency, and has demonstrated reliable effects both between groups and in individual score differences; prime-congruent semantic concepts may be activated in working memory and, rather than affective reactions, the valence of these semantic concepts guide the evaluations of the target, that is, the pictograph (Blaison et al., Emotion, 12, 403–412, 2012). Even if this semantic route is not the primary mediator, semantic processing is likely to be involved in evaluative responses, or at least, influence the processes leading to them (Storbeck and Clore, Cognition & Emotion, 21, 1212–1237, 2007). Here, we studied misattributions using the AMP by examining freely generated words, which were quantified using natural language processing to investigate whether the semantic content was influenced by the manipulations. We labeled this method Quantitative Semantic Misattribution Procedure (QSMP). More specifically, in the present study we measured implicit attitudes caused by Black and White faces (i.e., primes) on the semantic content of explicit verbal responses to Chinese characters. Since verbal responding relies on both highly automated implicit processes and consciously reportable explicit execution, quantitative semantics allows for studying both implicit and explicit processes. We showed that the semantic representation of participants’ verbal responses was significantly different for black and white primes in the un-warned but not the warned group and that the semantically expressed valence was equal for black and white face primes, whereas black target faces had higher semantic valence than white. Clearly, the QSMP can be used to study evaluative and non-evaluative influence of racial attitudes.