Farmer Cristan A, Kaat Aaron J, Mazurek Micah O, Lainhart Janet E, DeWitt Mary Beth, Cook Edwin H, Butter Eric M, Aman Michael G
1 Department of Psychology, The Ohio State University , Columbus, Ohio.
2 Department of Pediatrics and Developmental Neuroscience, National Institute of Mental Health , Bethesda, Maryland.
J Child Adolesc Psychopharmacol. 2016 Feb;26(1):10-8. doi: 10.1089/cap.2015.0098. Epub 2016 Jan 8.
The measurement of aggression in its different forms (e.g., physical and verbal) and functions (e.g., impulsive and instrumental) is given little attention in subjects with developmental disabilities (DD). In this study, we confirm the factor structure of the Children's Scale for Hostility and Aggression: Reactive/Proactive (C-SHARP) and demonstrate measurement invariance (consistent performance across clinical groups) between clinic-referred groups with and without autism spectrum disorder (ASD). We also provide evidence of the construct validity of the C-SHARP.
Caregivers provided C-SHARP, Child Behavior Checklist (CBCL), and Proactive/Reactive Rating Scale (PRRS) ratings for 644 children, adolescents, and young adults 2-21 years of age. Five types of measurement invariance were evaluated within a confirmatory factor analytic framework. Associations among the C-SHARP, CBCL, and PRRS were explored.
The factor structure of the C-SHARP had a good fit to the data from both groups, and strict measurement invariance between ASD and non-ASD groups was demonstrated (i.e., equivalent structure, factor loadings, item intercepts and residuals, and latent variance/covariance between groups). The C-SHARP Problem Scale was more strongly associated with CBCL Externalizing than with CBCL Internalizing, supporting its construct validity. Subjects classified with the PRRS as both Reactive and Proactive had significantly higher C-SHARP Proactive Scores than those classified as Reactive only, who were rated significantly higher than those classified by the PRRS as Neither Reactive nor Proactive. A similar pattern was observed for the C-SHARP Reactive Score.
This study provided evidence of the validity of the C-SHARP through confirmation of its factor structure and its relationship with more established scales. The demonstration of measurement invariance demonstrates that differences in C-SHARP factor scores were the result of differences in the construct rather than to error or unmeasured/nuisance variables. These data suggest that the C-SHARP is useful for quantifying subtypes of aggressive behavior in children, adolescents, and young adults with DD.
发育障碍（DD）患者很少受到对不同形式（如身体攻击和言语攻击）和功能（如冲动性攻击和工具性攻击）的攻击行为测量的关注。在本研究中，我们确认了儿童敌意与攻击量表：反应性/主动性（C-SHARP）的因子结构，并证明了有和没有自闭症谱系障碍（ASD）的临床转诊组之间的测量不变性（各临床组表现一致）。我们还提供了C-SHARP结构效度的证据。
照料者对644名2至21岁的儿童、青少年和青年进行了C-SHARP、儿童行为清单（CBCL）和主动性/反应性评定量表（PRRS）的评分。在验证性因素分析框架内评估了五种测量不变性类型。探讨了C-SHARP、CBCL和PRRS之间的关联。
C-SHARP的因子结构与两组数据拟合良好，并且证明了ASD组和非ASD组之间严格的测量不变性（即等效结构、因子载荷、项目截距和残差，以及组间潜在方差/协方差）。C-SHARP问题量表与CBCL外化问题的相关性比与CBCL内化问题的相关性更强，支持了其结构效度。被PRRS分类为反应性和主动性的受试者的C-SHARP主动性得分显著高于仅被分类为反应性的受试者，而后者的评分显著高于被PRRS分类为既非反应性也非主动性的受试者。C-SHARP反应性得分也观察到类似模式。
本研究通过确认C-SHARP的因子结构及其与更成熟量表的关系，提供了其效度的证据。测量不变性的证明表明，C-SHARP因子得分的差异是结构差异的结果，而不是误差或未测量/干扰变量的结果。这些数据表明，C-SHARP有助于量化患有DD的儿童、青少年和青年攻击行为的亚型。
