Abstract:

PBIS is a data driven, tiered framework for behavioral interventions, where all students receive some level of pro-social intervention, a much smaller number receive targeted, preventive training to reduce group level risks, and an even smaller group receive intensive, indicated intervention after the fact (Sprague et al, 2007). It has become the established best practice model for behavioral interventions in the United States and beyond (Sugai, 2007). One major challenge to scaling this model is ensuring fidelity to best psychosocial and instructional practices among teachers who, for the most part, have never been trained to deliver these kinds of interventions (Fixen et al, 2006; Backer, 2002). Another is providing culturally responsive interventions in schools where cultural diversity is the norm for students, and wide cultural gaps between teachers and at least some of their students is common. A third major challenge is finding ways to share behavioral data to drive decision making, without violating the rights to privacy of individual children and their families.

Technology offers promise for meeting these challenges. It can provide school wide information systems that provide a data driven framework, especially for assessment. It can automatically collect dosage data as part of computerized, intervention delivery. It can - but not necessarily does - ensure fidelity to both psychosocial and instructional best practices. Expert system technology can lock best psychosocial practice in a software program and deliver the most relevant set of those practices to each learner (Ray, 2010). Digital media and interactivity can provide to each learner a breadth of instructional modes that few teachers could equal. Multimedia content can build in a rich diversity of voices, images and experiences to help cross cultural gaps. And computers can collect, sort, guard, report, and analyze data, beyond even the most efficient human's capacity.

So what is to fear? An analysis of data from more than 5000 students across fifty schools indicates one technology-enabled, social-emotional learning intervention is effective for all three tiers of behavioral intervention, across cultural groups. Significant, positive, psychological (Koffman et al, 2008; Delong-Cotti, 2007), behavioral (Ray et al, 2008) and academic (Bass et al, 2008) outcomes have been substantiated in randomized controlled trials, as well as under quasi-experimental conditions. However, that same data indicates that student use of the learner-centric program routinely goes far beyond what program designers anticipated. It is weighted heavily toward private exploration of trauma-related topics, with dosage automatically collected. Theoretically, teachers understanding the traumatic context of individual students’ lives might provide a bridge toward more personalized academic and behavioral education. However, it just as likely might destroy for students who most need it, the sense of school as a sanctuary from difficult life circumstances. How can technology be designed to simultaneously maximize the protections for children’s privacy, and leverage specific data for well-informed decision-making? Are these even the most relevant questions? The decade long evolution of the underlying assumptions and design of a data management system for a computer-based, behavioral intervention system now used in more than 550 school districts in six countries will be presented for critique.

Keywords:

Behavioral interventions, student privacy, social-emotional learning.