Effects of the Lessons in Character English Language Arts Character Education Program on Behavior and Academic Outcomes
Study design
Designed as an experimental trial, the study runs from spring 2007 to spring 2010 in 50 volunteer California elementary schools with teachers of grades 2–5. Recruited using mass mailings and established WestEd marketing channels, schools are randomly assigned to treatment and control groups. The intervention involves a one-day training course for treatment teachers in year 1—who in turn are expected to incorporate between 19 and 24 supplementary lessons into their classroom instruction for two academic years. No Lessons in Character professional development activities or coaching occurs in year 2 of implementation. Two cohorts of elementary schools are participating. Cohort 1 implementation is taking place in 2007/08 and 2008/09, with teacher professional development and coaching in late summer and early fall of year 1 (2007). Cohort 2 schools were recruited in spring 2008, with implementation scheduled for 2008/09 and 2009/10. Both cohorts are pooled in the data analyses. Teachers in control schools, meanwhile, continue with their regular professional development activities and instructional practices.
The study sample consists of around 15,000 students in grades 2–5 in 50 public elementary schools in California—34 in the Los Angeles and San Diego Metropolitan Statistical Areas, 13 in northern California, and 3 in central California. Thirty schools are in a large city or urban fringe of a large city, six in a mid-size city (population of 25,000–250,000), and fourteen in a small town or rural area. Each school serves around 350 students, though nine have fewer than 150 students, and two have more than 600. Approximately 58 percent of the students are eligible for free or reduced-price lunch, 49 percent are Hispanic, 30 percent are English language learner students, and 33 percent are White (California Department of Education 2009).
With 25 schools per condition (treatment or control) and an expected average of at least 224 students in each school, the projected sample size is sufficient for detecting program impacts on student outcomes (0.17–0.23 standard deviation). Specifically, that sample size can detect an effect size equal to about two-fifths of a year of growth experienced by grade 2 students on norm-referenced standardized tests in reading (Hill et al. 2008).