Abstract: Classifying Changes to Preventive Interventions: A Comparison of Adaptation Taxonomies (Society for Prevention Research 24th Annual Meeting)

Schedule:

Thursday, June 2, 2016

Seacliff C (Hyatt Regency San Francisco)

* noted as presenting author

Joseph N Roscoe, MSW, Graduate Student, University of California, Berkeley, Berkeley, CA

Kelly Whitaker, PhD, postdoctoral fellow, University of Washington, Seattle, CA

B.K. Elizabeth Kim, PhD, Postdoctoral Scholar, University of California, Berkeley, Berkeley, CA

Valerie Shapiro, PhD, Assistant Professor, University of California, Berkeley, Berkeley, CA

Introduction: Evidence suggests that preventive interventions can promote social emotional competence and protect against behavioral and emotional disorders (Durlak et al., 2011). However, diffusion of preventive intervention is often compromised by implementation issues (Durlak & DuPre, 2008), and adaptation may be necessary to accommodate diverse service environments (Castro, Barrera, & Steiker, 2010). Scholars have suggested that adaptations be proactive and systematic, but in practice, adaptations are regularly made during implementation, contributing to concerns about erosion of fidelity and the dilution of outcomes (Elliott & Mihalic, 2004).

This study examines how three adaptation models retrospectively classify adaptations. An empirical model developed by Moore, Bumbarger, & Cooper (2013), uses a three-category taxonomy of adaptation fit, timing, and valence. Castro, Barrera, & Martinez (2004) propose a two-category taxonomy that classifies adaptation according to dimension and form. The Ecological Validity Model (EVM; Bernal, Bonilla, & Bellido, 1995) is an adaptation taxonomy that differentiates among 8 different types of intervention modifications.

Methods: TOOLBOX is a social emotional learning curriculum implemented in 11 elementary schools in the Bay Area during the 2015 academic year. Our study includes survey responses from 137 teachers, of whom 84% were female and 59% were white. Three open-ended survey questions addressed program adaptation, and teachers provided 90 distinct responses. Of these, 26 responses were general comments unrelated to adaptation. The remaining 64 responses were classified according the criteria specified in each model’s adaptation taxonomy.

Results: The Moore, et al. model classified 61% of responses using one or more categories, and 90% of responses related to adaptations for philosophical fit. The Castro, et al. model classified 72% of responses using one or more categories, and 63% of responses related to adaptation delivery. The EVM model classified 61% of responses, 59% of which related to adaptation methods.

Conclusions: Providers were more likely to explain adaptations as attempts to improve the philosophical fit of the intervention (e.g., adding cultural context). They were also more likely to describe adaptations to intervention delivery methods (e.g., incorporating drawing activities). The models performed reasonably well in classifying adaptations retrospectively, though each has room for improvement. None classified more than 72% of provider responses, and responses were generally not specific enough to map onto each category of the multi-category models. With revision, the models can more effectively classify adaptations in community settings, and help providers plan for systematic adaptation.

Abstract: Classifying Changes to Preventive Interventions: A Comparison of Adaptation Taxonomies (Society for Prevention Research 24th Annual Meeting)

333 Classifying Changes to Preventive Interventions: A Comparison of Adaptation Taxonomies