RESUMEN
The purpose of this study was to independently assess the best-fitting factor models of the Social, Academic, and Emotional Behavior Risk Screener (SAEBRS) student and teacher forms. To do this, we used previously published confirmatory factor analysis procedures (see von der Embse, Iaccarino, et al., 2017) in an attempt to replicate the factor structure. Unidimensional, correlated-factors, higher order, bifactor, and bifactor with correlated residuals models were assessed. The bifactor model yielded the best fit for the student, χ² = 286.58, p < .001, χ²/df = 1.91, RMSEA = .070, CFI = .839, TLI = .796, WRMR = 1.047, and teacher forms, χ² = 502.44, p < .001, χ²/df = 3.78, RMSEA = .095, CFI = .977, TLI = .971, WRMR = 1.193. Nonetheless, the majority of the fit statistics indicated an adequate fit for the student form. The SAEBRS Total Behavior score was found to have the greatest reliability for the student, ω = .77, ωH = .76, and teacher forms, ω = .93, ωH = .86, as well. Model, factor, and item-level indexes indicated mixed support for unidimensionality versus multidimensionality on student and teacher forms. Generally, it is implicated that the SAEBRS overall score was the soundest score for screening risk with the student and teacher forms. However, future investigations could consider a wider variety of methods to test competing factor structures. (PsycInfo Database Record (c) 2023 APA, all rights reserved).