Validation of the screen for child anxiety related emotional disorders-revised (SCARED-R), and a study on the relationship among temperamental traits, attentional bias and anxiety in children and adolescents

Abstract

Introduction: This study was designed to validate the Screen for Child Anxiety Related Emotional Disorder-Revised (SCARED-R) for use in the local context. Furthermore, different components of anxiety-related attentional bias were studied with emotional spatial cueing task (ESCT). This study also attempted to investigate the relationship among attentional bias, temperamental traits of Negative Affectivity (NA) and Effortful Control (EC), and anxiety.

Method: Subjects were recruited from P.4 to F.3 students of mainstream Chinese speaking schools. The SCARED-R and self-report measures of NA and EC were administered. 508 child/adolescent subjects (mean age=11.7 years, SD=1.80; 327 boys and 181 girls) and 312 of their parents (child’s mean age=11.50; SD=1.77; 196 boys and 116 girls) were included in the analysis. In Stage 2 of the study, subjects were recruited from those with SCARED-R scores at 85th percentile or above and 50th percentile or below. 34 pairs of child/adolescent and parent subjects (mean age=11.89 years; SD=1.90) were assessed with the Chinese version of the Diagnostic Interview for Children-Version 4, with which the Anxiety group (n=14) and Control group (n=20) were identified. All subjects completed the SCARED-R and Spence Children’s Anxiety Scale (SCAS). The child/adolescent subjects did ESCT, which presented Angry, happy and neutral faces for 17 ms, 500 ms and 1250 ms.

Results: The mean total SCARED-R scores were 31.94 (SD=19.51) and 22.96 (SD=15.35) for child/adolescent and parent subjects. Internal consistencies (Cronbach’s α=.95 and .94) and one-month test-retest reliability (Spearman’s rho=.72 and .79, both at p<.001) for child and parent total SCARED-R scores were satisfactory. Total SCARED-R scores correlated significantly with total SCAS scores, and were significantly different between the Anxiety and Control groups. These provided support for SCARED-R’s convergent validity. Results of exploratory and confirmatory factor analyses suggested that SCARED-R was multi-dimensional.

Facilitated attentional bias and disengagement difficulties were not related as predicted. Angry cues induced greater disengagement difficulties than happy cues. Compared with Control group, Anxiety group showed greater disengagement difficulties when Angry cues were presented at cue duration of 1250 ms.

A direct and significant relationship existed between anxiety with NA, and between anxiety and EC. NA and EC played their own unique roles in affecting anxiety but the variance accounted by EC in addition to NA was small. EC did not moderate the relationship between NA and anxiety, though a very small mediating effect between NA and anxiety was noted.

Conclusion: This study provides support for the satisfactory psychometric properties of the SCARED-R for local use, showing its good potential of being used as a screening instrument for anxiety disorders. Despite the small sample size, anxious children/adolescents were found to have disengagement difficulties from threat cue at cue duration of 1250 ms, and this could be a parameter to consider in future design of attentional bias modification training. The major limitation of this study is its convenience sampling of subjects from community and the low participation/consent rate. This limits generalizability of the results, impedes an investigation of the sensitivity and specificity of the scale, and poses great constraints on statistical power.