As we noted, one of the persistent disputes in personality psychology is between those who believe that self-reports or observer methods should hold priority in the field. The programmatic efforts of David Funder and his colleagues demonstrate that multiple methods bring multiple perspectives to our efforts to understand the behavioral manifestation of personality traits. For example, people judging the behaviors of others perceive different cues as more relevant to personality than the individuals themselves (Funder & Sneed, 1993).
In other studies, the usefulness of self- and observer ratings of personality have been tested across a variety of domains, including predicting behavior, emotions, and personal negativity. The key to testing the utility of different methods is separating the perspectives of self and observer from the criterion of interest. To do this, Kolar, Funder, and Colvin (1996) set up a study in which the participants provided self-report personality ratings, close acquaintances provided an additional set of personality ratings, and the behavior of participants was coded from videotaped interactions. Thus, the two sets of predictors and criteria did not suffer from methodological overlap. For behavior in a typical social setting, such as meeting a stranger or having a discussion, observer ratings tended to predict behavior better than self-reports (Kolar, Fun-
der, & Colvin, 1996). For example, the correlation between self-reported tendency to initiate humor and actual behavior of initiating humor was .09. In contrast, a composite of the rating of the participants' tendency to initiate humor of two close acquaintances correlated .23 (p <.05) with actual behavior. Clearly, what we believe to be a joke is not perceived by others to be funny, which might explain why more people don't laugh at our jokes.
Spain et al. (2000) used a similar design to both replicate Kolar et al. (1996) and extend the design to see if self-reports might be superior in specific settings, such as when one is predicting emotion rather than behavior. Consistent with expectations, self-reported personality ratings were more strongly related to experience sampling assessments of emotion than observer ratings of personality traits. This presumably derives from the fact that emotions are internal events that are not always shared with others as overt, visible behaviors. Their private nature makes them a natural target for self-reports rather than observer ratings. Interestingly, self-reported personality ratings did better than observer ratings of personality in predicting social interactions. For example, self-reported extraversión was correlated with demonstrating social skills, as judged by a set of trained raters, whereas a composite of acquaintance ratings was essentially uncorrelated with the same behavior. In fact, for extraversión, self-reports were twice as good as observer ratings of extraversión in predicting behaviors.
Clearly, based on this research alone, we cannot make any strong generalizations about the superiority of self-reports and observer ratings of personality. This is itself important, as it undermines claims that any given perspective is superior. Studies that actually use multiple methods arrive at more equivocal conclusions. This point is driven home conceptually in a review of the utility of psychological assessment (Meyer et al., 2001). In describing the importance of using multiple methods of personality assessment in clinical settings, Meyer et al. (2001) argued that each method affords a clinician, and by default a researcher, information that may not be strongly overlapping, yet still quite valid. That is to say that asking parents about a child's depression may not result in high agreement with the child's assessment (e.g., Lewis, 1999). Rather than seeing this as an indictment of either perspective or the construct of interest, we should use both of these perspectives and more (e.g., teachers, peers, siblings) to gauge the nature and progress of the phenomenon. For example, a child may have effectively hidden depression from his or her parents, but not hidden the same phenomena from his or her peers. The discrepancy itself may be both interesting and relevant to the experience of depression, as it might reflect alienation and disengagement from parents that might be a contributing factor to the depression.
The perspective that no single method holds priority extends to arguments against the use of projective measures (Dawes, 1994). For example, in our meta-analysis of the longitudinal consistency of personality trait measures (Roberts & DelVecchio, 2000), we found that projective measures of personality traits were as consistent as observer and self-report methods of personality assessment. Moreover, in particular cases, projective measures outperform other methods, such as in the assessment of dependency (Bornstein, 1999). This does not to provide a ringing endorsement for projective tests, as it is clear that specific projective tests and particular measures derived from projective tests do not demonstrate adequate reliability and validity (Lilienfeld, Wood, & Garb, 2000). Nonetheless, blanket statements that they should not be used are not warranted given the evidence.
The idea that perspectives that differ in terms of their hierarchical relationship to personality provide different, yet equally valid information was demonstrated nicely by a recent study of satisfaction with one's vacation (Wirtz, Kruger, Napa-Scollon, & Diener, 2003). In this study, participants rated how satisfied they thought they would be with an upcoming vacation. In addition, they completed an online assessment of their emotional experiences during the vacation using experience-sampling methods. A week later, they rated how satisfied they were with their vacation. Interestingly, anticipated and retrospective ratings of satisfaction were much higher than online ratings of satisfaction, indicating a slight disjoint between actual experience and higher-order evaluations of that same experience. Moreover, the different methods yielded different information. The online experiences were strong predictors of the retrospective ratings of satisfaction, which were in turn the most important predictor of wanting to go on a similar vacation in the future. The effect of actual experience on the desire to go on a similar vacation was entirely mediated by the higher-order generalizations about satisfaction, which indicates that the different methods yielded complementary information rather than redundant information. This study counters the argument that online assessments should be prioritized over broader, sometimes retrospective reports of personality (e.g., Kahneman, 1999), as it was the global self-reports that predicted long-term intentions rather than direct, behavioral measures of experience.
Within the trait domain, we find many of the classic arguments about multiple methods, such as the utility of self-reports versus other ratings and newer perspectives manifest in assessing personality across multiple levels of breadth or across different contexts. Consistent with the neosocioanalytic framework that a differentiated conceptualization of personality leads to a multimethod approach, each of these different methods revealed complementary and useful information. What we still lack, of course, are theoretical systems to account for the complementary rather than overlapping nature of the information gleaned from different methods. Systems like the RAM model are a step in the right direction, but more conceptual and theoretical work is needed.
Was this article helpful?