A response to Johnson and Hogan

Abstract

I will try to keep my response to Johnson and Hogan (1980) as brief as possible: The role-playing issue. Although test-taking (and personality in general) undoubtedly involves a certain amount of role-playing or selfpresentation, such role-playing can take two fundamentally different forms. On the one hand, an individual may attempt to present an image that is consistent over a variety of different situations or occasions. In this case the individual’s role or “face” (Goffman, 1967), including its particular enactment on a given personality test, carries many of the same implications that we associate with the concept of personality (e.g., consistency, individual differences). On the other hand, role playing may also refer to an individual’s attempt to alter his or her performance to fit the particular situation or audience at hand. In this second case, which is clearly the form of self-presentation or impression management that we were concerned with in Experiment 1 of the Meehan, Woll, and Abbott (1979) study, the individual’s role would appear to be inconsistent with any meaningful conception of personality and with any attempt to assess personality. Thus, our conclusion from Experiment 1 still stands: if subjects can manage their impressions on the SEA as readily and with as high a degree of consensus as our results indicate, it seems rather misleading to construe that scale as a measure of attitudes (or a disposition to adopt a set of attitudes-f. Hogan, 1973) or of personality. The social desirability issue. Johnson and Hogan argue that scores on the SEA cannot be accounted for by the social desirability of the test items because “these ratings” (presumably referring to the SDSV ratings) are context(or audience-) dependent, and because there are individual differences in such ratings. The point at issue, however, is not whether each subject’s score can be completely accounted for by his or her social