Key points are not available for this paper at this time.
The author extends his (1967) earlier study that (a) classified trait adjectives to unconfound evaluative and descriptive aspects, (b) made a selection from those that could be unconfounded, and (c) omitted those that could not. The present study adds representation of the achievement and affiliation adjectives omitted then. 10 undergraduates made inferences from 30 new traits (and 90 old ones) to 15 new scales (and 40 old ones). Factor analysis of the 120 traits produced 3 larger descriptive dimensions. Factor analysis of the 55 scales produced confounded factors, but these could be deliberately rotated to general evaluation, plus the descriptive dimensions. The dimensions corresponded well with those from interpersonal traits and psychopathology, but to only 3 of the Big Five factors derived from R. B. Cattell's (1947) 35 scales. The discrepancies are discussed as plausibly due to differences in representativeness or to differences between judgments about people and about traits (as opposed to claims for a close correspondence between them). (20 ref) (PsycINFO Database Record (c) 2013 APA, all rights reserved)
Dean Peabody (Wed,) studied this question.