Skip Navigation
Impacts of Comprehensive Teacher Induction:

NCEE 2009-4072
August 2009

Summary of Findings

The report presents findings from an experimental test of the impact of comprehensive teacher induction on student achievement in beginning teachers‘ classrooms and on the teachers‘ retention rates in urban elementary schools. In ten of the study districts, a comprehensive induction program was implemented during beginning teachers‘ first year in the classroom. In the remaining seven study districts, comprehensive induction was implemented during beginning teachers‘ first two years in the classroom. This design does not allow for and should not be used to make direct comparisons between the districts that received one year of treatment and districts that received two years of treatment, but instead allows us to investigate the effectiveness of one-year programs separately from that of twoyear programs. The main findings are summarized below.

  • During their first year in the classroom, in both one- and two-year districts, treatment and control teachers‘ reports showed statistically significant differences in the amount and types of support received. Treatment teachers were more likely than control teachers to report having an assigned mentor (90 versus 70 percent of teachers reported having an assigned mentor in one-year districts; 94 versus 79 percent in two-year districts) and reported spending more time per week with a mentor (87 versus 67 minutes in one-year districts; 124 versus 81 minutes in two-year districts). Treatment teachers reported spending more time being observed by mentors (34 versus 10 minutes during the most recent full week of teaching in one-year districts; 38 versus 17 minutes in twoyear districts) and meeting with mentors together with other first-year teachers (29 versus 9 minutes in one-year districts; 38 versus 11 minutes in two-year districts).
  • During their second year in the classroom, treatment teachers in one-year districts received less support than did control teachers. During Year 2, we found a statistically significant difference favoring the control group in teachers‘ likelihood of having an assigned mentor and in the amount of time teachers spent per week with a mentor. Treatment teachers were less likely than control teachers to report having an assigned mentor (20 versus 29 percent) and reported spending less time per week with a mentor (19 versus 39 minutes).
  • During their second year in the classroom, treatment teachers in two-year districts received more support than did control teachers. During Year 2, we found a statistically significant difference favoring the treatment group in teachers‘ likelihood of having an assigned mentor and in the amount of time teachers spent per week with a mentor. Treatment teachers were more likely than control teachers to report having an assigned mentor (80 versus 34 percent) and reported spending more time per week with a mentor (82 versus 48 minutes).
  • No impacts of comprehensive teacher induction were found on student achievement during teachers‘ second year in the classroom. In both one- and two-year districts, we did not find statistically significant impacts on student achievement across all elementary grade levels in reading or math during the teachers‘ second year.
  • No impacts of comprehensive teacher induction were found on teacher retention rates after two years. There was also no evidence that comprehensive teacher induction induced a change in the kind of teachers retained within the district. In both one- and two-year districts, we did not find statistically significant impacts of comprehensive teacher induction on teacher retention rates in the school, district or profession after two years. In both one- and twoyear districts, we did not find statistically significant impacts on the composition of the district teaching workforce after two years, whether measured by district stayers‘ impacts on student achievement or by their professional background characteristics (for example, SAT/ACT scores or whether the teacher attended a highly selective college).
  • In a correlational (nonexperimental) analysis of induction and student test scores, the relationship between four composite induction measures (considered jointly) and test scores was statistically insignificant for both math and reading. When we tested the variables individually, one of the four measures of beginning teacher support (years had a mentor) was positively related to math scores (coefficient = 0.12, p-value = 0.015) and none were related to student achievement in reading. The significant result can be interpreted as a student scoring 12 percent of a standard deviation higher on the math test for each year the beginning teacher had a mentor. The nonexperimental results should be interpreted with caution because the analyses are correlational and not causal.
  • In the correlational analysis of induction and teacher mobility, there was a positive relationship between the four composite induction measures and retention that was statistically significant for both retention in the district (pvalue= 0.016) and retention in the profession (p-value=0.001). When we tested the induction indices one at a time, one of the four explanatory variables was positively related to retention in the district, none were positively related to retention in the profession, and none were negatively related to either type of teacher retention. The estimate of the regression coefficient on the Induction Services Index for remaining in the district was 0.02. This implies that, for example, if the retention rate in a district were 80 percent, then an additional induction service, such as meeting with a study group in one semester, would be associated with a district retention rate of 82 percent, all else equal. As mentioned above, the nonexperimental results should always be interpreted with caution because the analyses are correlational and not causal.

Top