An empirical investigation of issues in the assessment of social skill

Abstract

Research in the assessment and training of social skills has been hampered by the absence of adequate measurement and assessment instruments. The present study investigated a number of assessment issues in the area of social skill. These issues concerned the reliability and validity of untrained judges, the utility of role play, the behavioural consistency of skilled and unskilled subjects and the predictive power of nonverbal, paraverbal and verbal behaviours in social skill. Forty-three subjects were videotaped in two roleplays and a waiting interaction with a confederate of the opposite sex. Nine untrained judges provided criterion ratings of global social skill, while raters scored the paraverbal taped performances on a and verbal measures. The results indicated that untrained judges were fairly reliable; behaviour in role play and waiting interaction differed for unskilled subjects, but not for skilled subjects; skilled subjects did not show greater behavioural variability than unskilled subjects. A series of regression analyses revealed that the amount and timing of speech had the greatest influence on skill judgements. In general, predictive behaviours were able to account for considerably more of the variance in unskilled subjects than in skilled subjects. These findings suggest that nonverbal and paraverbal behaviours are significant at the lower end of the skill spectrum, but that other factors, probably verbal, influence judgements of high skill. The implications of these results for social skill assessment and training and for further research in the area are discussed. Show more