Classifying Changes to Preventive Interventions: Applying Adaptation Taxonomies

• Published in: The Journal of primary prevention Vol. 40; no. 1; pp. 89 - 109
• Main Authors: Roscoe, Joseph N., Shapiro, Valerie B., Whitaker, Kelly, Kim, B. K. Elizabeth
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.02.2019; Springer Nature B.V
• Subjects: Child psychology; Coding; Community and Environmental Psychology; Elementary schools; Health Psychology; Intervention; Medicine; Medicine & Public Health; Online Surveys; Original Paper; Public Health; Social Emotional Learning; Social skills; Taxonomy; Teachers; Measurement; Prevention; Social and emotional learning; Implementation; Adaptation
• ISSN: 0278-095X; 1573-6547
• DOI: 10.1007/s10935-018-00531-2

High-quality implementation is important for preventive intervention effectiveness. Although this implies fidelity to a practice model, some adaptation may be inevitable or even advantageous in routine practice settings. In order to organize the study of adaptation and its effect on intervention outcomes, scholars have proposed various adaptation taxonomies. This paper examines how four published taxonomies retrospectively classify adaptations: the Ecological Validity Framework (EVF; Bernal et al. in J Abnorm Child Psychol 23(1):67–82, 1995 ), the Hybrid Prevention Program Model (HPPM; Castro et al. in Prev Sci 5(1):41–45, 2004 . https://doi.org/10.1023/B:PREV.0000013980.12412.cd ), the Moore et al. (J Prim Prev 34(3):147–161, 2013 . https://doi.org/10.1007/s10935-013-0303-6 ) taxonomy, and the Stirman et al. (Implement Sci 8:65, 2013 . https://doi.org/10.1186/1748-5908-8-65 ) taxonomy. We used these taxonomies to classify teacher-reported adaptations made during the implementation of TOOLBOX™, a social emotional learning program implemented in 11 elementary schools during the 2014–2015 academic year. Post-implementation, 271 teachers and staff responded to an online survey that included questions about adaptation, yielding 98 adaptation descriptions provided by 42 respondents. Four raters used each taxonomy to try to classify these descriptions. We assessed the extent to which raters agreed they could classify the descriptions using each taxonomy ( coverage ), as well as the extent to which raters agreed on the subcategory they assigned ( clarity ). Results indicated variance among taxonomies, and tensions between the ideals of coverage and clarity emerged. Further studies of adaptation taxonomies as coding instruments may improve their performance, helping scholars more consistently assess adaptations and their effects on preventive intervention outcomes.