Metatraits: Evidence relevant to the validity of the construct and its implications.

Further evidence for the metatrait construct was obtained by comparing correlations among theoretically related variables for Ss traited on both constructs, untraited on both constructs, or traited on 1 construct and untraited on another. Using hierarchical multiple regression, Ss traited on both constructs evidenced a substantially higher correlation between the 2 constructs than Ss untraited on 1 or both constructs. These findings replicated across 3 theoretical relationships. Study 2 showed that the effects of traitedness are not a function of the certainty or importance with which Ss view the traits being correlated but are hypothesized to represent the presence vs. absence of a trait structure in personality. The implications of the results for construct validation are discussed. The idea that attributes differ in their relevance to an individual's self-concept has intuitive appeal. However, researchers in psychology have often implicitly assumed that individuals are capable of denning themselves on all theoretically derived constructs. A further assumption has been that the score an individual receives on a particular scale is more important than the relevance of that measured attribute to the individual's selfconcept. Both assumptions have been challenged by researchers emphasizing the differential relevance with which attributes are related to personality and self (Allport, 1937; Baumeister, 1991; Baumeister Bern Kenrick Lamiell, 1981; Markus, 1977). These researchers have noted that all traits are not equally applicable to all individuals. For example, although two individuals may receive the same score on an extraversion scale, the trait of extraversion may be highly relevant to one person but not to the other. The strength and relevance with which people possess particular traits, as well as their perceived variability on a trait, moderates the correlation between scores on a given