Findings from Experiemntal Analyses
Students of AC teachers did not perform statistically differently from students of TC teachers. Although average differences in reading and math were generally negative, they were not statistically significant, as shown in Exhibit 5.
In addition to estimating the effects on student achievement of having a high- or lowcoursework AC teacher, we examined effects within several subgroups to determine whether differences in teachers' effectiveness occurred within other dimensions even though differences did not exist overall. Specifically, we examined the relative effects of teachers in subgroups defined by state, current coursework status, grade level, and teaching experience.
All AC teachers in California were from high-coursework programs, and they accounted for half of all high-coursework AC teachers in the sample. Students of AC teachers in California scored lower on math than students of their TC counterparts, and the effect size (–0.13) was statistically significant. The effects of high-coursework AC teachers in other states was small (–0.01) and not statistically significant.
Students of AC teachers who were taking courses during the study year, toward either teacher certification or an advanced degree, had lower math scores than students of their TC counterparts (effect size = –0.09). The effect in reading was not statistically significant. Furthermore, neither the effect on reading nor the effect on math scores was significant for students of AC teachers who were not taking coursework during the study year.
We found no evidence that AC teachers had a different effect on their students' math or reading achievement for different grade levels. There were no statistically significant differences between the lower elementary grades (K to 1) and the upper ones (2 to 5) for either the high- or the low-coursework AC teachers.
We found no evidence that students of AC teachers with less experience (1 to 2 years) had statistically significant different math or reading achievement, relative to their TC counterparts, than those with more experience (3 to 4 or 5 or more years). The one statistically significant difference pertained to students of low-coursework AC teachers in their third or fourth year of teaching, whose students scored lower in reading and math than students of their TC counterparts. Inferences based on these findings should be made with caution because the subgroup sizes were small and the experience levels of the TC comparison teachers varied.
With a single exception, ratings of classroom practices measuring the instruction received by students of AC and TC teachers did not differ. We found no statistically significant differences in VCOT scores between low-coursework AC teachers and their TC counterparts in the quality of their literacy and math instruction, as shown in Exhibit 6. High-coursework AC teachers also scored no differently from their TC counterparts on five of six VCOT measures, but they scored lower (by 0.40 SD) on the classroom culture dimension in teaching literacy, and the difference was statistically significant.