The judgments of personality that are made both by psychologists and by "real people" provide a focus for my research. The research particularly concerns the circumstances under which such judgments are likely to be accurate.

This interest in accuracy naturally leads me to be concerned with the principles and techniques of personality assessment. Some of my research has concerned the construction and evaluation of various assessment techniques, including self-report measures of personality and judgments gathered from people who know well the persons who are judged. I have also written about the "person-situation controversy" within personality psychology, which has concerned the question of whether personality traits exist to a degree strong enough to merit their inclusion within psychological theory and practice. The latest direction of research in our lab concerns the psychological assessment of situations. While person variables and, to a lesser extent, behavioral variables have been an important part of psychological research, too little attention has been paid to identifying the important dimensions of situations. Supported by an NSF grant, we are developing the Riverside Situational Q-sort (RSQ) and testing its applicability to understanding the three-way interaction among persons, situations, and behaviors.

The other part of my research program examines the processes by which people come to make judgments of the personalities of themselves and each other in daily life. My focus is upon the various factors that make such judgments more and less likely to be accurate. I have divided these factors into four categories, which I call "good judge," the possibility that some people are better judges of personality than are others; "good target," the possibility that some people are easier to judge than others; "good trait," the possiblity that some traits are easier to judge accurately than others; and "good information," the possibility that more or just better information leads to enhanced accuracy in personality judgment. These potential moderators of accuracy are being examined within a large data set that includes the self-judgments, friends' judgments, and videotaped samples of the behavior of a large sample of subjects.