Last weekend I had the pleasure of joining Dr. Jie Zhang of WKU in presenting our recent research examining the relationship between teacher perceptions and the implementation of Positive Behavior Interventions and Supports (PBIS) at the annual conference of the American Education Research Association (AERA) held in Philadelphia.
Co-authors on our study included WKU EdD student Chunling Niu, GRREC behavior consultant Dr. Kelly Davis (a graduate of WKU's doctoral program), and former WKU colleague Dr. Kyong Hee Chon.
Our study explored whether teachers in schools implementing Positive Behavior Interventions and Supports had better preceptions of their working conditions than teachers in non-PBIS schools. PBIS provides a framework for establishing school-wide behavioral expectations and engaging teachers in regular analysis of discipline data to set goals, monitor progress, and trouble-shoot problems. The goals of PBIS include enhanced school culture and climate and a focus on positive student behavior (rather than a strictly punitive approach to student conduct).
To explore teacher perceptions in PBIS and non-PBIS schools, we analyzed data from the 2011 TELL (Teaching, Empowering, Leading, and Learning) Kentucky survey, administered to teachers across the Commonwealth by the Kentucky Department of Education. TELL measures teacher perceptions of their working conditions on eight constructs, including Managing Student Conduct (a concept closely related to the goals and processes of PBIS).
We used MANOVA and ANOVA statistical techniques to compare TELL responses between 151 PBIS-implementing schools and 144 demographically-matched non-PBIS schools. As predicted, teachers in PBIS schools reported significantly better perceptions of Managing Student Conduct than teachers in non-PBIS schools, but also had better perceptions of school leadership. This suggests that, at least in terms of its impact on teacher perceptions, PBIS might serve to improve working conditions in schools.
We were further able to divide PBIS schools based on the extent to which they faithfully implemented key structures of PBIS, creating three groups (high-fidelity, medium-fidelity, and low-fidelity implementers). Among these groups, high-fidelity schools had significantly higher perceptions of Managing Student Conduct, Community Support and Involvement, and Teacher Leadership than their lower-fidelity counterparts, suggesting that if schools plan to embrace PBIS, they should work hard to ensure that all aspects of the program are fully implemented in order to achieve maximum results.
In a final analysis, we used hierarchical linear regression to examine linkages between PBIS implementation and student achievement using 2011 Kentucky student outcome data. We found no significant differences between PBIS and non-PBIS schools in this regard, although high-fidelity PBIS schools had significantly higher achievement levels than low-fidelity implementers. This finding should be viewed with some consideration of the study's limitations, however. High-fidelity implementers may have already had higher student outcomes and these results may bear no actual linkage to PBIS implementation.
We appreciated the feedback we received from conference participants. Many complimented the inclusion of the fidelity of implementation variable, which is often not considered in program evaluation studies.
We believe this study makes an important contribution in the growing literature about PBIS and also about the TELL Kentucky survey (which was administered again in 2013 and has become a mandated component of school improvement efforts in the state). Previous analyses have found relatively little linkage between teacher perceptions as measured by TELL and student outcomes. In other words, TELL does not distinguish well between high and low-performing schools. But the application of TELL for program evaluation purposes, as utilized in this study, perhaps holds more promise.
We will seek to publish results of this research in a peer-reviewed journal in coming months. In the meantime if you would like more information about the study, please e-mail me.