Project Activities
This project has four specific aims:
- Determine the long-term impact of the Tier 1 + 2 supports relative to Tier 1 only on middle and high school behavioral and academic outcomes such as disruptive behavior and academic achievement; juvenile justice outcomes such as arrests and detainment; postsecondary outcomes such as college enrollment, persistence, and degree attainment; and early labor outcomes such as workforce participation and early labor market earnings.
- Determine whether the long-term impacts are moderated by student risk based on baseline teacher behavioral rating scores and baseline academic achievement or whether Tiers 1 + 2 are most effective (or close gaps) for particular student groups including students receiving special education services, students of color, disadvantaged students, and English language learners.
- Examine the generalizability of program impacts by estimating these Tiers 1 + 2 effects in the population of students and schools in the state of Maryland.
- Determine the long-term cost-benefit of Tiers 1 + 2, including the return on investment of the impact across the education, juvenile justice, and workforce sectors.
Structured Abstract
Setting
The original PBISplus trial included 42 Maryland public, comprehensive elementary schools. All schools in the trial were previously trained to implement PBIS Tier 1 by the PBIS Maryland State Team (MSDE and Sheppard Pratt Health System). In this follow-up study, the researchers will link trial data with archival data from all students and schools in Maryland through the MLDS in order to conduct a long-term follow-up, draw propensity score matched schools for exploratory quasi-experimental contrasts with non-PBIS schools, conduct generalizability analyses, and perform a cost-benefit analysis across multiple service sectors.
Sample
In the previous randomized control trial (RCT) study, 29,569 students and from 3,202 staff participated. The student sample were majority male (52 percent) and not White (54 percent African American, 8 percent Hispanic of any race, and 33 percent White). Most staff were female (91 percent) and White (75 percent). The schools served a sizable proportion of students who received free and reduced meals (43 percent) and were highly mobile.
The original PBISplus RCT examined the combined impacts of Tier 1 + 2 PBIS supports as compared to Tier 1 only. PBISplus provided training and coaching in Tier 2 supports, including function-based thinking, a simplified functional behavioral analysis model, Check-In/Check-Out, and culturally responsive practices to reduce disproportionality in discipline and special education referrals.
Research design and methods
The initial efficacy trial examined the efficacy of the PBIS Tiers 1 + 2 relative to Tier 1 only, using a group RCT design. The researchers randomized elementary schools to receive either PBISplus (Tiers 1 + 2; intervention; n = 22 schools) or continued participation in the statewide standard PBIS Tier 1 programing (comparison; n = 20). The researchers collected four waves of data (baseline and three spring timepoints) in the original RCT. In this follow-up study, the research team will merge data from the RCT with MLDS data across multiple service sectors spanning the years 2007-08 through 2022-23 to support this 16-year follow-up study.
Control condition
The comparison schools only received training in PBIS Tier 1 supports.
Key measures
RCT data include teacher-ratings of students' behavior, engagement ,and social emotional functioning; academic performance; staff self-reports of efficacy, behavior management, and school climate; and fidelity of implementation. MLDS data include standardized test scores, high school completion, educational service utilization, discipline problems and arrests, postsecondary readiness, postsecondary entry and degree attainment, and labor market outcomes.
Data analytic strategy
The researchers will leverage intent-to-treat and moderated main-effects models using multi-level modeling and multilevel growth models to examine primary aims. They will use complier average causal effect estimation and propensity scores for a quasi-experimental exploratory sub aim. They will use propensity scores in a separate generalizability analysis.
Cost analysis strategy
The research team will use the ingredients method to enumerate the costs of the PBISplus intervention based on information derived from fidelity data and coaching log data, taking into consideration opportunity costs. They will also conduct a cost-benefits analysis using a shadow pricing approach to estimate the cost savings of PBISplus in relation to any long-term benefits identified across multiple service sectors (education, labor market, justice system).
People and institutions involved
IES program contact(s)
Project contributors
Products and publications
Products: The research team will disseminate the main, moderated, generalized, and cost effects of PBIS Tiers 1 + 2 through peer-reviewed manuscripts, presentations, cost trainings, and free online research reports/briefs working in close collaboration with practice partners and stakeholders from several Maryland agencies and organizations including http://www.pbismaryland.org/resources/research-briefs and http://mldscenter.maryalnd.gov/CenterOutput.html).
Related projects
Supplemental information
Co-Principal Investigators: Bradshaw, Catherine; Pas, Elise; Goldstein, Ross
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.
