The many faces of assumed similarity in perceptions of personality

We propose that most coefficients of assumed similarity in acquaintance ratings are based on actual similarity between observer and target rather than any bias to project self onto other. We assessed assumed similarity, actual similarity, and rater accuracy in 62 university roommate dyads who rated each other on 74 personality-related variables. Our findings indicate that our roommates (a) were actually alike on many variables, (b) knew this, rather than assumed it, and (c) applied this knowledge to enhance the accuracy of their ratings. We outline a model of judgment that urges three distinct concepts: true assumed similarity, known similarity, and spurious similarity. We conclude that personality ratings among acquaintances reflecting true assumed similarity are probably rare.     

Big five factors and facets and the prediction of behavior

The authors compared the Big 5 factors of personality with the facets or traits of personality that constitute those factors on their ability to predict 40 behavior criteria. Both the broad factors and the narrow facets predicted substantial numbers of criteria, but the latter did noticeably better in that regard, even when the number of facet predictors was limited to the number of factor predictors. Moreover, the criterion variance accounted for by the personality facets often included large portions not predicted by the personality factors. The narrow facets, therefore, were able to substantially increase the maximum prediction achieved by the broad factors. The results of this study are interpreted as supporting a more detailed approach to personality assessment, one that goes beyond the measurement of the Big 5 factors alone.     

What is the central feature of extraversion? Social attention versus reward sensitivity

R. E. Lucas, E. Diener, A. Grob, E. M. Suh, and L. Shao (2000) recently argued that the core of the personality dimension of Extraversion is not sociability but a construct called reward sensitivity. This article accepts their argument that the mere preference for social interaction is not the central element of Extraversion. However, it claims that the real core of the Extraversion factor is the tendency to behave in ways that attract social attention. Data from a sample of 200 respondents were used to test the 2 hypotheses with comparisons of measures of reward sensitivity and social attention in terms of their saturation with the common variance of Extraversion measures. The results clearly showed that social attention, not reward sensitivity, represents the central feature of Extraversion.     

Facial Features as Personality Cues

Two studies examined the effects of the appearance of specific facial features on attributions of personality. In Study 1, photographs of men and women were computer-manipulated to have larger than average or smaller than average eye size, and wider than average or narrower than average eye spacing. In Study 2, eye size and mouth fullness were similarly altered. Although it was found that neither eye spacing nor mouth fullness had any effect on perceptions of the targets' personality or physical characteristics, eye size had strong effects in both studies. Analyses of covariance revealed that the personality trait ratings that varied with eye size were mediated primarily by perceived differences in the targets' masculinity-femininity and babyfacedness, and to a lesser extent by attractiveness.     

The validity of formal and informal personality assessments

Two recent articles by M. Burisch (1984, American Psychologist, 37, 214–227; 1984, Journal of Research in Personality, 18, 81–98) have presented data bearing on the traditional methods by which personality is assessed. Based on a review of selected empirical studies and analyses of new data, Burisch has concluded that (a) single-item global self-ratings generally yield greater coefficients of criterion validity than questionnaire measures of the same traits, and (b) most existing personality questionnaires can be shortened considerably without any consequent loss in validity. These conclusions stand in direct contrast to previous appeals by Epstein and others for the use of extensive aggregate measures of personality derived according to the axioms of psychometric theory. In response to Burisch, we describe measurement artifacts and strategies of test item selection that account for some of his conclusions and we reiterate the need for the application of conventional construct-based questionnaires in the assessment of personality.     

The reliability of aggregated measurements: Lessons to be learned from psychometric theory

In a recent article, H. D. Day, D. Marshall, B. Hamilton, and J. Christy (1983, Journal of Research in Personality, 17, 97–109) have reported on a series of simulation studies designed to investigate properties of various coefficients of behavior stability as applied to cross-situational data. The results of their research have prompted them to suggest precautions regarding the measurement of behavioral regularity with conventional indices of response consistency. A critical review of their paper reveals aspects of the data, procedures, and interpretations of Day et al. (1983) which undermine many of their conclusions. Notable problems include (a) inappropriate comparisons of internal consistency reliability coefficients based on nonparallel analyses, (b) anomalous data which may be due to errors of rounding or recording, and (c) the failure to recognize certain assumptions underlying reliability estimation. More importantly, it is shown that the investigation of Day et al. (1983) represents primarily an empirical verification of several measurement principles for which algebraic proofs are amply documented.     

Optimizing the validity of personality assessments: The importance of aggregation and item content

The predictive validity of a psychological measure can be improved by minimizing measurement errors through increases in the length of the assessment (aggregation) and, for an assessment of finite length, by making use of objective strategies for choosing from all available component measures. Two prominent considerations in selecting individual measures to be aggregated involve standards of (a) item content (construct approach) and (b) item/criterion association (empirical approach). Personality trait scales of different lengths were assembled for this study in order to represent features of the construct and empirical methods of selection. It was observed that (a) although reliability and validity generally increased with test length, aggregation beyond a certain point can fail to be expedient; and (b) although the prediction performance of empirically derived measures initially surpassed that of construct based assessments, the superiority of the empirical scales did not generalize to trait criteria that were not used as a basis for item selection. The data are interpreted as providing support for a theory-based program of test development where substantive considerations involving item content play a major role. The findings are also viewed as encouragement for conventional conceptualizations about organized dimensions of behavior.     

On ad hoc personality scales: A reply to Burisch

Two recent articles by M. Burisch (1984, American Psychologist, 37, 214–227; 1984, Journal of Research in Personality, 18, 81–98) have presented data bearing on the traditional methods by which personality is assessed. Based on a review of selected empirical studies and analyses of new data, Burisch has concluded that (a) single-item global self-ratings generally yield greater coefficients of criterion validity than questionnaire measures of the same traits, and (b) most existing personality questionnaires can be shortened considerably without any consequent loss in validity. These conclusions stand in direct contrast to previous appeals by Epstein and others for the use of extensive aggregate measures of personality derived according to the axioms of psychometric theory. In response to Burisch, we describe measurement artifacts and strategies of test item selection that account for some of his conclusions and we reiterate the need for the application of conventional construct-based questionnaires in the assessment of personality.     

On the accuracy of ratings of personality by strangers

Past studies in person perception have shown unexpectedly high correlations between targets' self-ratings of personality traits and ratings of those targets made by stranger judges. Possible reasons for the apparent accuracy of strangers' ratings are examined, including personality judgments that are based on (a) the perceived social desirability of targets, (b) known sex-linked base rates of behavior, and (c) the inevitable presence of valid personality cues in the typical stranger rating study. A computer simulation study is then described that illustrates that a relatively small number of nonrandom accurate judgments among many random judgments can produce substantial overall correlations between self-ratings and stranger ratings. The data of the study provide some clarification of past empirical results of both stranger and acquaintance ratings of personality.