This study examined the ecological validity of role-played dating interactions. Forty-five male undergraduates, categorized into low-, medium-, and high-frequency dating groups, participated in four heterosexual social (heterosocial) situations in a laboratory. Two situations were role plays simulating dating interactions, and two were more naturalistic waiting-period interactions. Controls for confederate familiarity and role-play order were used. Judges rated subjects as more skillful in role plays than waiting periods and rated low'-frequency daters as more anxious than high-frequency daters. High-frequency daters rated themselves as more skillful than did low-frequency daters. Subjects rated the waiting-period interactions as more like "real life" and their behavior in waiting periods as more representative of everyday heterosocial interactions. However, analyses of variance did not result in major differences between role plays and waiting periods, and correlational analyses indicated that the relative ranking of subjects was similar in role plays and waiting periods. Implications for the use of roleplay assessments in the heterosocial skill area are discussed. Curran (1977) and Hersen and Bellack (1977) have recently pointed out numerous methodological flaws in heterosexual social (heterosocial) anxiety and skill research, with a major and pervasive weakness being the lack of validity data on many of the assessment instruments. The data that are available provide inconsistent support for any given measure's validity. Because appropriate evaluation of selection and treatment procedures relies on the reliability and validity of assessment techniques (Paul, 1969), the lack of basic psychometric data on the assessment procedures seriously limits the validity of research in this area.Many investigators have used role plays of various common interactions to assess heterosocial anxiety and skill; for example, sitting in a pizza parlor waiting for a pizza (Curran, 1975), initiating a conversation Requests for reprints should be sent to

Two studies were conducted in order to obtain evidence regarding the external validity of a social simulation measure of social skills. The first study employed a contrasted group design, with the groups consisting of National Guardsmen and psychiatric patients. As predicted, trained judges rated the performances of the National Guardsmen in the simulations as more skillful and appearing less anxious than the psychiatric patients. In the second study, trained judges" ratings of psychiatric patients" performances in the simulations were found to be significantly related to ratings made by interviewers who had conducted a structured social history interview with the patients, nurses' ratings based on their observations of patients after 3 days on the psychiatric unit, a research assistant's ratings based on contact with the patient during the experiment, and finally, the subject's own selfratings. In discussing the findings, it is hypothesized that one reason why support was found for the external validity of the simulation strategy in this study, in contrast to some previously reported studies, was the use of molar ratings in this study as opposed to the more molecular rating procedures used in those previous studies.

Groups of high socially anxious, low socially skilled (socially incompetent), and low socially anxious, high socially skilled (socially competent) under-graduate males were selected using both a self-report measure of social anxiety and behavioral ratings of both skill and anxiety in a simulated role-play of a dating situation. Selected subjects were then requested to (1) listen to a ten-minute videotape of a female speaker discussing her life and (2) press a switch whenever they felt a response would communicate understanding or rapport. The hypothesis that socially incompetent and competent individuals would respond with the same frequency but would differ in the timing or distribution of their responses was affirmed. The results suggested timing of responses to be an important component of social skill and called into question the adequacy of using frequency counts of behaviors to investigate the specific nature of social skill. Implications for both assessment and treatment of social skill deficits are offered.

The generalizability of social-skills and social-anxiety global ratings made by judges trained in six different experimental laboratories was examined. The stimulus material consisted of videotapes of psychiatric patients and of a control sample of National Guardsmen interacting in two types of social simulation typically ernployed in social-skills assessment: brief role plays and more extended interactions. Moderate degrees of generalizability across laboratories were found for the social-skills and social-anxiety ratings based on the brief role plays. For the more extended interaction, moderate generalizability was obtained for the anxiety ratings, with some relative differences found among laboratories for the skills ratings. The results were viewed as encouraging, since numerous mitigating factors worked against establishing strong generalizability across laboratories. Implications for social-skills assessment and treatment are discussed.One of the most common procedures employed in assessing social skills involves engaging subjects in laboratory-based simulated social interactions. A subject's performance in the simulated interaction is then evaluated by trained judges. Not only is this one of the most common procedures but, asThis study was funded in part by a research grant from the Veterans Administration.