Are Implicit Attitudes “Implicit” for their Holders? Testing GATA Construct Validity

This article examines the construct validity of the “Graphical Associative Test of Attitudes” (GATA). Presumably, its measurand is implicit attitude or implicit fraction (or implicit component) of holistic attitude, which depends on a theoretical conceptualization. A commonly accepted characteristic of implicit phenomena of the mind is their inaccessibility, or at least difficulty, to introspection. Some later studies have claimed that respondents are able to predict their results on the Implicit Associations Test, which in this case means that implicit content is accessible. Leaving aside the discussion about the degree of accessibility of implicit phenomena, we used the general theoretical assumption and operational design of some of our colleagues to test the construct validity of our own “indirect” measurement method. Would our respondents be able to predict their personal scores on GATA? The results of the experiment have shown — no, they would not. The respondents' predictions could be explained far more reliably by their explicit attitudes and by their expectations about the attitudes of the social environment. Thus, we have not found any refutation of the initial assumption that the GATA measurand is an implicit fraction of attitudes. This allows us to focus our further research on the problem of the accuracy of the GATA measure.