Cotter Katie L, Evans Caroline B R, Smokowski Paul R
1 Arizona State University, Phoenix, AZ, USA.
2 University of North Carolina at Chapel Hill, NC, USA.
J Interpers Violence. 2017 Apr;32(7):1087-1102. doi: 10.1177/0886260515587665. Epub 2016 Jul 11.
Measures of violent behavior are often assumed to function identically across different groups (e.g., gender, race/ethnicity). However, failure to verify measurement invariance can lead to biased cross-group comparisons. The current study examines the measurement invariance of the Violent Behavior Checklist-Modified across genders and race/ethnicities. Using multiple group confirmatory factor analysis, configural and metric invariance are assessed in a sample of racially/ethnically diverse middle and high school students ( N = 4,128) in two rural counties. Results indicate that the Violent Behavior Checklist-Modified has partial measurement invariance across genders and race/ethnicities. Specifically, four out of six items were non-invariant across genders, while one out of six items was non-invariant across race/ethnicities. Findings suggest that the latent factor of violence may be qualitatively different across males and females. Implications are discussed.
暴力行为的测量方法通常被假定在不同群体（如性别、种族/民族）中具有相同的功能。然而，未能验证测量不变性可能导致有偏差的跨群体比较。本研究考察了修订后的暴力行为清单在不同性别和种族/民族中的测量不变性。使用多组验证性因素分析，在两个农村县的种族/民族多样化的初中和高中学生样本（N = 4128）中评估构型不变性和度量不变性。结果表明，修订后的暴力行为清单在性别和种族/民族方面具有部分测量不变性。具体而言，六个项目中有四个在不同性别间不具有不变性，而六个项目中有一个在不同种族/民族间不具有不变性。研究结果表明，暴力的潜在因素在男性和女性之间可能存在质的差异。文中讨论了相关影响。
