This study examined the relationship of teacher variables and growth in student mathematics achievement, with a focus on elementary students of color and poverty. Student achievement data was linked to 156 third through sixth grade teachers in an urban Washington State school district. Measures of Academic Progress (MAP) growth scores for 3,558 students were calculated as a percentage of expected growth (PEG) and aggregated to the teacher level to form the dependent variable, the classroom percentage of expected growth achieved (CPEG).

Of four teacher variables studied, only advanced degrees (r = .32, p <.01) and experience (r = .17, p <.05) were significantly related to student achievement. In a multiple regression analysis, advanced degrees explained 9.7% of student achievement variance beyond demographic variables, but experience did not significantly add to R² beyond demographics and advanced degrees. College math course taking and college selectivity were positively related to student achievement, but not significant. No interactions between variables were detected.

Although measurable teacher characteristics explain little of between classroom variance, this study reinforces findings that teachers matter greatly. The range of achievement between classrooms one standard deviation above and below the mean is equivalent to a half year of expected growth. High performing classrooms score above expected growth, regardless of the diversity level of the classroom. Finally, evidence is presented that an equitable distribution of teachers may be able to narrow achievement gaps.