Replicating the CGI Experiment in Diverse Environments
Co-Principal Investigators: Walter Secada and Juli Dixon
Purpose: Cognitively Guided Instruction (CGI) is a widely used professional development program for mathematics instruction in elementary school. Researchers propose to replicate a previous efficacy study to examine the impact of this intervention on student achievement in mathematics when implemented with a larger and more diverse sample of students. The study also seeks to identify the principal, teacher, and student characteristics that moderate intervention impact and to explore the mechanisms through which the intervention affects student mathematics achievement. Results from this study will describe the ideal conditions for CGI implementation and CGI-related work in other mathematical domains (e.g., algebra and fractions).
Project Activities: Researchers will test the efficacy of CGI using a cluster-randomized controlled trial research design. Twenty-two elementary schools will be randomized to the treatment or control condition. The impact of the intervention will be tested each year using teacher knowledge measures and a teacher collaboration survey before the summer institutes, researcher observation of teacher instruction throughout the year, and student end-of-year standardized achievement tests. Three cohorts of students will be followed through third-grade. Researchers will examine the impact of the intervention on teacher and student outcomes, as well as the moderating effects of teacher and student characteristics and the mediating role of teacher knowledge and instructional practice on the intervention's impact on student outcomes.
Products: The products of this project will be evidence of the efficacy of the Cognitively Guided Instruction (CGI) teacher professional development program for first- and second-grade teachers and their students. Peer-reviewed publications will also be produced.
Setting: Participating elementary schools are located in a school district in Florida, consisting of urban, suburban, and rural communities.
Sample: Researchers will recruit a total of 22 elementary schools, serving students of varying racial/ethnic backgrounds, English language learner (ELL) status, and reduced/free lunch status. From these schools, researchers will recruit a total of 264 teachers (132 first-grade and 132 second-grade) and all their students during their involvement in the project (approximately 18 students per classroom for three cohorts, totaling 7,128).
Intervention: Cognitively Guided Instruction (CGI) is a widely used professional development program for mathematics instruction in elementary school. Teachers are taught to use a categorization scheme for types of math problems and for intuitive student thinking about math problems. This two-year program consists of two summer institutes and two sets of follow-up days throughout the school year. In the summer institutes, videos illustrating student thinking and the CGI book (Children's Mathematics: Cognitively Guided Instruction) are used to introduce teachers to CGI. Then, teachers practice writing problems for particular populations. During the school year, teachers interview students, discuss student work and thinking, and apply knowledge from summer institutes.
Research Design & Methods: Researchers will randomly assign schools to the treatment or control condition. Teachers in the treatment condition will participate in the two-year CGI professional development program. The impact of the intervention will be tested each year using teacher knowledge measures and a teacher collaboration survey before the summer institutes, researcher observation of teacher instruction throughout the year, and student end-of-year standardized achievement tests. Three cohorts of students will be followed through third-grade. One student cohort will have had two years of instruction from a CGI-trained teacher (i.e., first- and second-grade) and will be followed for three years. Another student cohort will have had one year from a CGI-trained teacher (second-grade) during the teacher's first year in the intervention and will be followed for two years. Another student cohort will have had one year from a CGI-trained teacher (second-grade) during the teacher's second year in the intervention and will be followed for one year.
Before the start of the intervention, researchers will analyze state archival data and pre-test teacher knowledge data of all participating teachers to verify baseline equivalence on teacher knowledge, collaboration measures, and adequacy of random assignment procedures. Teachers in the treatment condition will attend the first CGI summer institute. At the end of the summer institute, the teachers will complete post-test teacher knowledge measures. These teachers will receive follow-up support and researchers will collect observational data throughout the academic year. The second year of the intervention will consist of another summer institute, another year of follow-up support, and researcher observation. Researchers will collect the final measures of teacher knowledge and collaboration at the end of the two-year intervention.
Control Condition: In the control condition, teachers received practice-as-usual plus a 2-day information session on Common Core State Standards in mathematics.
Key Measures: Teacher pedagogical content knowledge will be measured using three separate instruments—one assessing mathematical knowledge of teaching, another assessing knowledge of the CGI framework, and another assessing teachers' knowledge of student thinking processes. Teacher collaboration will be measured using a survey of their perceptions of school culture, instructional practice, and team-level instructional practice. Teacher instructional practice will be measured using classroom observations rated for the teachers' engagement with students' mathematical thinking.
Student mathematics achievement will be measured using the Iowa Test of Basic Skills and the Florida Comprehensive Assessment Test. In addition, students will complete measures of content knowledge specific to the CGI framework and a word problems assessment. Researchers will also collect information about student demographic characteristics (i.e., race/ethnicity, ELL status, and socioeconomic status) and prior academic achievement.
Data Analytic Strategy: Researchers will conduct two- and three-level model analyses to examine the impact of the intervention on teacher and student outcomes, including student achievement after varying years instructed by a CGI-trained teacher and varying years since instructed by a CGI-trained teacher. Researchers will also test the moderation effects of teacher characteristics (i.e., baseline knowledge and collaboration) and student characteristics (i.e., race/ethnicity, ELL status, socioeconomic status and prior academic achievement) and explore the role of teacher knowledge and instructional practice as mediators of the intervention's impact on student outcomes.
Journal article, monograph, or newsletter
Schoen, R.C., Bray, W., Wolfe, C.M., Tazaz, A.M., and Nielsen, L. (2017). Developing An Assessment Instrument To Measure Early Elementary Teachers' Mathematical Knowledge For Teaching. The Elementary School Journal, 118(1): 55–81.
Whitacre, I., Schoen, R.C., Champagne, Z., and Goddard, A. (2016). Relational Thinking: What's the Difference?. Teaching Children Mathematics, 23(5): 302–308.