The Predictive Power of Teacher Practice in Explaining Student Growth

In recent years, states across the nation have increased their interest to develop specific teacher accountability measures and improve student achievement. On August 6, 2012, the state of New Jersey approved the TEACHNJ Act, which reformed tenure laws and linked student growth to a teacher’s evaluation. The ultimate goal of the TEACHNJ Act is to “raise student achievement by improving instruction through the adoption of evaluations that provide specific feedback to educators” (TEACHNJ Guide). The 2013-14 school year was the first full year of implementation, which included student growth percentile (SGP) scores as one component of a teacher’s evaluation. The purpose of this quantitative study was to examine the relationship between teacher practice and student growth. The study determined the probability that a student will have typical or high growth on the state assessment in relation to their teacher’s practice score based on classroom observations. Some of the essential questions regarding this research are as follows: Are teacher-level variables such as gender, ethnic background, and age significant predictors of student growth? Are school-level variables such as school performance status (Comprehensive schools, Target schools, and NonStatus schools) and percent of student subgroup ethnic composition significant predictors of student growth? How is student growth in language arts and mathematics impacted by a teacher’s effectiveness as measured by the practice score received, when one controls for teacher- and school-level characteristics? To what extent does the relationship between teacher effectiveness and student outcome vary from year two of AchieveNJ to year five of AchieveNJ? The sample population for the 2014-2015 school year will consist of 1132 students (n = 1132) with a valid language arts SGP and 1087 students (n = 1087) with a valid mathematics SGP. The sample population for the 2017-2018 school year consisted of 1484 students (n = 1484) with a valid language arts SGP and 1473 students (n = 1473) with a valid mathematics SGP. The study involved 12 to 14 schools with different grade configurations, performance status, and student ethnic composition. This study was a cross-sectional explanatory design in which logistic and hierarchical logistic regression methods were used to test the relationships between the dependent variable (student growth) and independent variables (teacher characteristics, school characteristics, teacher practice). The design consisted of three separate models used to answer four research questions. A logistic regression analysis will be used to analyze Model 1 (teacher characteristics on student growth) and Model 2 (school characteristics on student growth). In Model 3, a hierarchical logistic regression analysis was used to better interpret the impact of teacher practice and teacher and school characteristics on student growth. Research question four compared and analyzed the significant findings between Model 3 in the 2014-2015 school year and Model 3 in the 2017-2018 school year. This study will provide insight for educational leaders and policymakers on the positive relationship between teacher practice and student growth and recommends that this type of research continue to explore how other variables influence student learning growth based on how teachers deliver instruction