Doing Dosage Differently: Exploring Indicators of “Strategy” Implementation

Introduction: Success of effective preventive interventions in routine settings is driven by the strength of implementation (Durlak & DuPre, 2008). However, implementation suffers if sub-optimal amounts of intervention are delivered. Inadequate dosage may be of particular concern in school settings, where lesson-based interventions can impinge on other curricular obligations. Some argue that the infusion of intervention elements into routine practices (i.e., a strategy-based) may be less of an implementation challenge than a lesson-based dose form (Embry & Biglan, 2008; Jones & Bouffard, 2008). However, dose form (i.e., method for dispensing active ingredients; Warren, Fey, & Yoder, 2007) is rarely reported in studies (Zeng, Law, & Lindsay, 2012).

The present study examines lesson-based and strategy-based dose forms during a routine implementation of TOOLBOX, a social-emotional learning (SEL) intervention. The TOOLBOX Implementation Research Project (TIRP) featured two intervention conditions: “Standard” (a fully manualized lesson-based approach with accompanying infusion strategies) and “Primer” (greatly abbreviated lessons with an emphasis on infusion strategies). The study explores 1) how dosage differs by condition, and 2) how dosage changes over time.

Methods: TOOLBOX was implemented in four elementary schools during the 2015-16 academic year. Classroom teachers (n=95) participated in an online survey on three occasions. The sample was 95% female, 62% European, 11% Hispanic, and 7% African American. Based on school-level randomization, 46 teachers implemented TOOLBOX Standard and 49 implemented TOOLBOX Primer. Growth models were used to assess dosage over time.

Results: Teachers reported similar levels of lesson-based dosage (e.g., frequency, duration of lessons) across conditions at all time points. Similarly, there were no significant differences in strategy-based dosage (e.g., frequency of using TOOLBOX in classroom rituals) across conditions at any time, except that teachers in the Standard condition wore their fan-decks (i.e., program lanyards with visual representations of the tools) more often than teachers in the Primer condition on all occasions.

On average, lesson-based dosage (e.g., time spent teaching favorite lesson: β=0.37, p<0.001), increased over time in both conditions. Most indicators of strategy-based dosage, on average, decreased over time in both conditions. One strategy-based indicator (time spent reinforcing TOOLBOX: β= 0.25, p= 0.001) increased over time.

Implications: Emphasis on different dose forms did not consistently change teacher implementation behavior. Implementation monitoring, however, should include multiple dose forms since lesson-based dosage and strategy-based dosage had different patterns over time.