Journal of Abnormal Child Psychology, Vol. 26, No. 3, 1998, pp. 213-220 Cross-Informant Agreement in the Assessment of Social Phobia in Youth Patricia Marten DiBartolo1,5 Anne Marie Albano2 David H. Barlow,3 and Richard G. Heimberg4 The use of multiple informants in the assess- ment of anxious children is a widely recommended practice. A number of studies have examined the re- liability of assessment of anxiety disorders in chil- dren, using parents as an important source of information regarding the child's symptomatology (e.g., Rapee, Barrett, Dadds, & Evans, 1994; Silver- man & Nelles, 1988). Despite the fact that clinicians demonstrate acceptable agreement in the assignment of anxiety disorder diagnoses, parent-child agree- ment often tends to be poor for the anxiety disorders and other internalizing disorders (Edelbrock, Costello, Dulcan, Conover, & Kalas, 1986; Rapee et 1Department of Psychology, Smith College, Northampton, Massa- chusetts. 2Department of Psychology, University of Louisville, Louisville, Kentucky. 3The Center for Anxiety and Related Disorders, Department of Psychology, Boston University, Boston, Massachusetts. 4Department of Psychology, Temple University, Philadelphia, Pennsylvania. 5Address all correspondence to Patricia M. DiBartolo, Smith Col- lege, Clark Science Center, Northampton, Massachusetts 01063. al., 1994). Previous investigations have generally found anxious children to report a larger number of more intense symptoms than their parents (cf. Bird, Gould, & Staghezza, 1992; Edelbrock et al., 1986). Recent data suggest that the age of the child may have a significant impact on parent-child agree- ment regarding the diagnosis of social phobia in the child. Rapee et al. (1994) reported that informant variance was reduced in children over the age of 10; that is, parents and older children were more likely to agree on the presence of social phobia. When as- sessing younger children (<10 years), Rapee et al. noted that the children often minimized their social anxiety relative to the reports of their parents. Indeed, the validity of child self-report may be uncertain when the principal diagnosis is social pho- bia. Given the social concerns and fear of negative evaluation that are definitional of social phobia, the inclusion of parent report may increase the validity of the diagnostic assessment. Thus, one goal of this investigation included examining parent-child symp- tom level agreement on measures of social fear and 213 0091-0627/98/0600-0213$15.00/0 © 1998 Plenum Publishing Corporation In the present study involving children and adolescents with a principal diagnosis of social phobia, we measured parent-child agreement regarding social anxiety symptoms. Addition- ally, we examined variables related to the severity of the children's social phobia symptoms as reported by children and as rated by clinicians. Examination of cross-informant agreement indicated little difference between mean parent and children ratings of the children's social fears. In contrast, there was a significant difference in parent and children ratings of the children's avoidance, with parents endorsing greater degrees of social avoidance. Children's report of social avoidance was negatively related to scores on a measure of self-presentational concerns (i.e., social desirability). Clinicians' determination of the severity of the children's social phobia was also influenced by the children's self-presentation as well as parent report of social avoidance and children's depression scores. Thus, this differential weighting by the clinician of parent versus child report may be related to the finding that children's self-re- ported social avoidance was negatively related to their concerns regarding positive self-pres- entation. Results suggest the need to consider the impact of social desirability when examining clinical characteristics of children and adolescents with social phobia. KEY WORDS: Social phobia; children; adolescents; cross-informant agreement; social desirability.