| Since No Child Left Behind has become law, there has been increased use of data to drive instructional decisionmaking for academics. However, the focus placed on improving behavior and using data to drive decisions for school-wide behavior improvement has not occurred with the same fervor. Behavior needs often are identified too late in a student's school career, which limits intervention efficacy and increases cost. This study tests the reliability and validity of the School-wide Efficient Behavior Screening (SWEBS); (Pierson, 2003) and discusses how educators used the SWEBS to make instructional decisions to sort students into three levels of differentiated interventions.; The SWEBS was administered in three elementary schools in a large midwest town (urban) to determine interrater reliability. Items with poor interrater reliability were eliminated and a new factor analysis revealed three factors anticipated a priori: Academic Survival Skills, Conduct, and Social-Emotional Needs. Test-retest reliability was found to be .921 and interrater reliability was .939. Validity was found to be satisfactory, as the SWEBS identified students already identified to be at-risk through other means. Teachers rated the behaviors of 591 students. Thirty-six of the teachers participated in focus groups to interpret the data, and were surveyed later about the experience. Qualitative findings showed that the SWEBS items seem to identify students appropriately with social, emotional, and behavioral needs. Teachers also reported that it was useful for designing and monitoring school-wide interventions, determine systemic needs in the school improvement process, and design instructional groups. Implications and directions for future research are discussed. |
