Comparability of self- and other-rated personality structure

It is commonly accepted that gathering information via multiple assessment methods (e.g., interview and questionnaire, self- and informant report) is important for establishing construct validity. Although numerous articles report convergent and discriminant agreement correlations between self- and other ratings of personality, studies of the structure of personality from such ratings are less common. The present study addresses this gap using a meta-analytic data set (N range = 157-9,295) of various versions (i.e., self- and other-report, full-length and short alternative format) of the Schedule for Nonadaptive and Adaptive Personality (SNAP; Clark, 1993; Clark, Simms, Wu, & Casillas, 2014). We hypothesized that (a) structures across all measure formats would be highly comparable and (b) to the extent that they were dissimilar, perspective (self vs. other) and measure format (long vs. short form), respectively, would influence comparability. Results revealed strong congruence among 3-factor structures (Negative Emotionality, Positive Emotionality, and Disinhibition vs. Constraint) across all versions of the SNAP, suggesting that personality as assessed by this broad measure of personality traits across the normal-abnormal spectrum has a robust structure across different rater perspectives and rating formats. Because the comparability analyses were highly congruent and differences among the comparisons were minimal, we concluded-contrary to our expectations-that different formats and different rater perspectives have little effect on structural comparability. Results generally support Funder's (1995) realistic accuracy model, suggesting that trait relevance, cue detection, and information usage are key factors in structuring informant ratings. Limitations of the present study and implications for future research are discussed. (PsycINFO Database Record (c) 2019 APA, all rights reserved).