A Randomized, Wait-List Controlled Effectiveness Trial Assessing School-Wide Positive Behavior Support in Elementary Schools

Elementary schools in Hawaii and Illinois, randomised to immediate training versus wait list, over 3 years

US elementary schools; state personnel delivered SWPBIS training and technical assistance.

The foundational randomised trial for school-wide PBIS, and it is developer-led: Horner and Sugai are the framework's originators and directors of the federally funded PBIS Technical Assistance Center. The primary demonstrated link is training to implementation fidelity, which is a compliance outcome rather than a child outcome. The authors are explicit that office discipline referrals could not be interpreted causally because experimental control was absent for that variable.

Training and technical assistance were functionally related to improved implementation of universal SWPBIS practices, and improved implementation was functionally related to improvements in the perceived safety of the school setting and in the proportion of third graders meeting or exceeding state reading standards. Office discipline referral levels were comparatively low, but the authors state that the absence of experimental control for that variable precludes inference about the impact of SWPBIS on it.

Low as designed (school randomisation); the live threat is developer involvement and outcome selection, not heredity.

The Maryland trials (Bradshaw et al.) reproduce the discipline-referral and suspension reductions on a larger randomised sample, so the behavioural claim has support. The reading claim here rests on a single mediated analysis within a small wait-list trial and has not been reproduced by a well-powered independent achievement evaluation.