Three determinants of the factor structures of personality traits are investigated. The 1st, selection of variables, was controlled by using 57 bipolar scales, selected to be representative of common trait terms. In analyses of 7 data sets, variants of the "Big Five" factors were always found. Factor similarities were very strong for the 3 largest factors, which were transformed into general evaluation and 2 descriptive dimensions. As a 2nd determinant, judgments about real people were compared with judgments about the conceptual relations among traits. Factor structures based on the 2 types of judgments are similar, but those based on conceptual judgments tend to be simpler. The 3rd determinant involved the degree of restriction of the sample to evaluatively homogeneous targets. Restriction of range reduced the size of all factors, especially Factor II. Findings from previous studies are integrated within this framework.

The common emphasis on the importance of evaluation in judgment was critically examined in a study of inferences between traits. Sets of traits were selected to remove the usual confounding between evaluative and descriptive aspects of judgment. Subjects made inferences from 90 traits to 40 scales denned by two opposing traits. Results revealed the secondary importance of evaluation. On 70 items where they were directly opposed, the descriptive aspects were always decisive over evaluation. Factor analysis showed that none of the factors was evaluative. General descriptive dimensions suggested by factor analysis could account for the evaluative consistency of inferences in general. Evidence suggested that evaluation is typically based on a descriptive judgment of the degree of extremeness. A model was proposed for the systematic relation of evaluation to descriptive judgment.We normally make judgments of people pects of judgments. The present paper will using trait terms such as "kind" or "rash." critically examine some recent forms of this What are the important aspects of such emphasis. These come from two main sources: judgments? Some aspects seem of obvious im-1, The factor analysis of semantic differenportance: these judgments typically involve tial judgments. Osgood, Suci, and Tannenus in liking or disliking the person and in baum (1957) analyzed the correlations beevaluating him as relatively good or bad. tween judgments on a variety of adjective Many interpretations emphasize the great im-scales. In several analyses the first and portance of such affective or evaluative as-largest factor was regularly interpreted as i This investigation was supported in part by Public evaluative. It was claimed that evaluation had Health Service Research Grant MH-06094 from the a "dominant role [p. 38]" as the "first de-National Institute of Mental Health, The author's terminant operating in meaningful judgment debt to others is even greater than usual. The study rp 4^1 » was stimulated by discussions and joint pilot work ' _L' . , , . , , , with S. E. Asch. Over several years, many students-2. Theories of the consistency of evaluation unfortunately too many to list here-collaborated in or affect. Logically simpler than the factor the work. Martin Ostwald and Richard Brandt, analysis of many judgments is the emphasis colleagues in classics and philosophy, contributed evaluative or affective consistency betheir broader viewpoints. L. R. Goldberg and N. S.•' Thompson gave important advice. Alan Coller and tween as few as two judgments. The most Ann Severson served as assistants on the project, general examples of this emphasis are such At various points, the project benefited from the baknce theories as Heider (19S8) Q d computer facilities at Haverford College, Swarthmore •, m i_ Vi T_ j College, the Health Science Institute in Los Angeles, and lannenbaum (1955), Rosenberg and and the

A strong argument has been made for specifying the possible characteristics of personality by using trait adjectives from the natural language. This argument implies a rationale for the selection of variables to measure personality characteristics: One should select variables to be representative of trait adjectives in the natural language. In this study I argue that this rationale has never been carried through, and I make an attempt to do so. This requires defining a pool of trait adjectives, classifying them according to similarity of meaning, and selecting variables to represent this classification. Previous classifications have either led to no selection of variables or to a selection that is not representative. A pool of 571 terms was taken from three existing lists. These terms were classified into categories, and scales were selected to represent the categories. A first application of these scales was then made. Judges made ratings of similarity of meaning on 57 scales—each defined by two opposite trait-adjectives—for the corresponding 114 single terms. Factor analysis of the scales produced six factors. These showed some resemblance to the well-known Big Five factors, but with major differences. Instead of five factors of comparable size, the first three factors were predominant, representing 70% of the total variance. The Culture factor appeared as Intelligence; Emotional Stability as a small, only partly differentiated, factor. These differences may reflect both (a) differences between judgments about traits (internal structure) and judgments about people (external structure), and (b) representativeness (e.g., the previous nonrepresentation of intelligence and overrepresentation of Emotional Stability).

The psycholexical approach to personality structure in American English has led to the Big Five factors. The present study considers whether this result is similar or different in other languages. Instead of placing the usual emphasis on quantitative indices, this study examines the substantive nature of the factors. Six studies in European languages were used to develop a taxonomy of content categories. The English translations of the relevant terms were then classified under this taxonomy. The results support the generality of Big Five Factor III (Conscientiousness). Factors IV (Emotional Stability) and V (Intellect) generally did not cohere. Factors I (Extraversion) and II (Agreeableness) tended to split when this was necessary to produce 5 factors. The analysis was extended to several additional studies.

No abstract