Discriminant Validity of the Screen For Anxiety Related Emotional Disorders (SCARED): A Methodological Replication Study


Marlane Cully

Defense Date


Graduation Date

Fall 1-1-2003



Submission Type


Degree Name



Counselor Education and Supervision (ExCES)


School of Education

Committee Chair

Lisa Lopez Levers

Committee Member

Boris Birmaher

Committee Member

Emma C. Mosley

Committee Member

Jeff A. Bridge


Anxiety Disorders and Children, Discriminate Validity, SCARED


Child and adolescent anxiety disorders form the most common group of psychopathological disorders effecting youth. These disorders are associated with depression, suicide, substance abuse, psychosocial problems, school problems and low self-esteem (Curry and Murphy, 1995; Keller et al., 1992; Klein, 1994; Lipsitz et al., 1994; Reinherz et al., 1993). Yet anxiety disorders are often unrecognized, underdiagnosed and misdiagnosed. This in part is due to assessment instrumentation in the area of youth anxiety, which historically have had various measurement problems. Self-report scales have often been recognized as global scales, limited in their ability to discriminate between anxiety versus depression, anxiety verses disruptive disorders and children with a respective diagnosis, panic disorder, generalized anxiety disorder, separation anxiety disorder, social phobia and those children with other anxiety diagnosis but without that disorder. However, previous research has found support for the discriminant validity of one self-report scale, the Screen for Anxiety Related Emotional Disorders (Birmaher et al., 1996; Birmaher et al., 1999; Muris et al., 1997).

This study is a methodological replication study. The aim of this study is to replicate previous findings on the discriminant validity of the SCARED by incorporating a new sample of children between the ages of 5 and 19, and their parents. There were one hundred and eighty nine (n=189) children and parents who attended a mood/anxiety clinic in this new sample. Five hundred and thirty one subjects (n=531) had previously been reported on (Birmaher et al., 1996 and Birmaher et al., 1997) bringing the current total of the sample to seven hundred and twenty (n=720). The receiver operator characteristic (ROC) curve method was used to determine specificity and sensitivity of the specific anxiety disorders. An analysis of variance (ANOVA) was used to compare means across the groups, anxiety and depression, anxiety and disruptive disorder and children with a respective anxiety diagnosis, GAD, SAD, social phobia, specific phobia (school) separation anxiety disorder but without the disorder. Both methods were utilized in determining the discriminant validity of the SCARED. Results: The t test statistically demonstrated that the total child anxiety score and each of the 5 factors significantly discriminated between anxious and disruptive children (p .05). Both the total child SCARED score (p .001) and the total parent SCARED score (p .004), except the social anxiety factor, discriminated between the anxious children and depressed children. The total child SCARED score (p .05) and the total parent SCARED score (panic/somatic, GAD, SAD, and social phobia factors significantly discriminated between those children with the respective diagnosis (i.e., panic disorder, GAD, SAD and social phobia) and those children with other anxiety diagnoses but without the given disorder (all noted comparisons p .008). Only the parent school avoidance factor discriminated between those children with school phobia and those without school phobia but with other anxiety disorders. The SCARED shows promise as a screening instrument for anxiety disorders. Implications for further research is evaluation of the discriminant validity of the SCARED in nonreferred samples.





This document is currently not available here.