1. Assessing the Psychometric Characteristics of a Child/Adolescent Behavioral and Emotional Risk Self-Report Screener across Grade Levels and School Districts
- Author
-
Society for Research on Educational Effectiveness (SREE), Harrell-Williams, L. M., Dowdy, E., and Twyford, J.
- Abstract
The BESS is a relatively new screening system for identifying behavior and emotional risk (BER) in children and adolescents. Psychometric evidence regarding this instrument is important for researchers and practitioners considering the use of the BESS for identifying BER in students. This study seeks to provide and evaluate evidence for the use of the Student form of the BESS. The following objectives guided this study: (1) assess the internal structure of the BESS Student Form using factor analysis; (2) determine if the items perform differently across two locations with different ethnic characteristics via Rasch-based DIF analysis; and (3) assess the odds of elevated risk levels based on student demographics. Data were collected in two locations: the Los Angeles United School District (LAUSD), and Bibb County (Georgia). The first sample was collected from the Los Angeles United School District (LAUSD) as part of a larger IES funded research project investigating the longitudinal validity for screening (Advancing Children's and Teacher's Success through Early Screening and Intervention; U.S. Department of Education), and consisted of 273 elementary and middle school students. The second sample contained the responses of 4,074 students from 3 LAUSD high schools in the Northern region of the school district. The third sample was collected as part of a pilot for district-wide screening initiative in Bibb County. There were 1,874 students across the four schools in grades 9 through 12. To evaluate the constructs being measured by the BESS as well as evaluate the predictive validity for future educational outcomes, a longitudinal, trend survey approach was used. To evaluate the appropriateness of the BESS in high school-aged students, a cross-sectional survey approach was used. Data was collected at one time point for all high schools. Results show that the theorized four factor structure of the BESS holds for all school levels (elementary/middle and high school), no evidence of differential item functioning was found across the 30 items when comparing the LAUSD high school group to the Bibb high school group, and logistic regression analysis revealed that gender and location did not affect risk. Tables and figures are appended.
- Published
- 2013