Beyond the Basics: Exploring Fidelity of Red Light Purple Light! and Links with Child Outcomes

• In Event: Unpacking the Black Box of Early Interventions through Broadened Measurement of Implementation Fidelity

Abstract

Interventions aimed at improving children’s self-regulation and engagement in the preschool classroom are plentiful (e.g., Tools of the Mind, PATHS) and demonstrate a variety of effects on classroom and child outcomes (e.g., Diamond et al., 2019; Domitrovich et al., 2007). Red Light Purple Light! (RLPL) is a cost-effective, teacher-implemented and classroom-based intervention that demonstrates improved self-regulation and math in preschool children (Li et al., 2020; McClelland et al., 2019). Regarding fidelity, teachers report strong adherence to the RLPL program (McClelland et al., 2019), but the nuances in implementation—and their possible associations with child outcomes, have not been explored in detail. Using teacher-report, videotaped intervention sessions, and direct child assessments, this paper aims to describe RLPL implementation quality (i.e., fidelity, quality, and responsiveness; Berkel et al., 2011) and then explore associations to child outcomes.
Data come from 149 RLPL sessions (from 10 Head Start classrooms; six teachers) and 120 children who participated in an initial efficacy trial of RLPL (see Table 1 for child-level descriptives). Fidelity of implementation was measured with teacher daily logs and videotaped sessions across the RLPL sessions (for more about RLPL, see McClelland et al., 2019). In the daily log, which was completed after each session, teachers reported on three items: two related to quality (e.g., “I liked playing the games today) and one for responsiveness (i.e., children liked playing the games today”) using a 5-point Likert scale (strongly disagree to strongly agree). Average scores were created across sessions. Videotaped intervention sessions (n = 37) were double coded using a reliable measure that assesses implementation quality (4 items), responsiveness (1 item), and fidelity (3 items), all scored on a 3-point Likert scale (Tominey et al., 2019). Consensus scores were averaged across sessions. Child direct assessments of self-regulation (HTKS-R [McClelland et al., 2021] and Day-Night Stroop task [Gerstadt et al., 1994]), early numeracy/math (PENS; Purpura et al., 2015), and literacy (WJ letter-word identification; Woodcock et al., 2001) were conducted in English or Spanish at post intervention.
Overall, implementation quality was high across participants (Table 2). All of the implementation items displayed high means with limited variability, with a few notable exceptions (e.g., teacher incorporates children’s ideas and child engagement displayed more variability; Table 2). Correlations (Table 2) reveal teachers modeled actions more often when they reported feeling prepared and teachers enjoyed the games more when they were observed to display more positive classroom management techniques. Teachers reported a positive association between the manual being helpful and liking the RLPL games. With regard to child outcomes, teacher preparation (observed and teacher-report) and the teacher’s comfort explaining rules showed positive associations with children’s letter word identification (r = .25 [teacher preparation]) and early numeracy knowledge (r = .25 [explains rules]). Results suggest that teacher responsiveness, teaching quality, and fidelity within the intervention may relate to development of numeracy and literacy skills for children. Follow-up analyses include hierarchical regression models predicting child outcomes as well as exploring how associations may differ by RLPL conditions.

Authors

• Bridget E Hatfield, Oregon State University - Corvallis
  Presenting Author

• Christopher Gonzales, University of California - Davis

• Shauna L Tominey, Oregon State University - Corvallis

• Alexis Tracy

• Megan M McClelland, Oregon State University - Corvallis