Skip to Main Content
Table 3.
Attrition and Missing Data
Pooled Year 1Cohort 1 Follow-Up Year
Control MeanTreatment — Control Difference (SE)Control MeanTreatment — Control Difference (SE)
Administrator outcomes     
End-of-year survey 0.815 0.110**   
  (0.051)   
Post-conference survey 0.849 0.073   
  (0.053)   
Time use survey 0.962 0.033   
  (0.022)   
Teacher outcomes     
End-of-year survey 0.888 0.040   
  (0.031)   
Post-conference survey 0.877 0.038   
  (0.038)   
Official observation score 0.788 0.058   
  (0.054)   
Student outcomes     
End-of-year survey: Teacher has any students with data 0.893 0.021 0.875 −0.181** 
  (0.034)  (0.068) 
End-of-year survey: Share of student surveys returned 0.755 0.039 0.758 −0.069 
  (0.043)  (0.069) 
Test Scores: Teacher has any students with data 0.936 0.030 0.652 0.125** 
  (0.032)  (0.050) 
Test Scores: Share of students who are present 0.970 −0.010 0.859 0.033 
that are in analysis sample  (0.006)  (0.036) 
Pooled Year 1Cohort 1 Follow-Up Year
Control MeanTreatment — Control Difference (SE)Control MeanTreatment — Control Difference (SE)
Administrator outcomes     
End-of-year survey 0.815 0.110**   
  (0.051)   
Post-conference survey 0.849 0.073   
  (0.053)   
Time use survey 0.962 0.033   
  (0.022)   
Teacher outcomes     
End-of-year survey 0.888 0.040   
  (0.031)   
Post-conference survey 0.877 0.038   
  (0.038)   
Official observation score 0.788 0.058   
  (0.054)   
Student outcomes     
End-of-year survey: Teacher has any students with data 0.893 0.021 0.875 −0.181** 
  (0.034)  (0.068) 
End-of-year survey: Share of student surveys returned 0.755 0.039 0.758 −0.069 
  (0.043)  (0.069) 
Test Scores: Teacher has any students with data 0.936 0.030 0.652 0.125** 
  (0.032)  (0.050) 
Test Scores: Share of students who are present 0.970 −0.010 0.859 0.033 
that are in analysis sample  (0.006)  (0.036) 

Notes: The adjusted difference between control and treatment is the result of a regression of the dependent variable against fixed effects for randomization strata and a treatment indicator. Standard errors (SE) are reported in parentheses, and allow for clustering within-school. For student-level outcomes, samples reflect students of teachers for whom we have available data. Students whose teachers did not provide class rosters for the end-of-year survey or were not in test-score files, were not included.

**Significant at the 95% level.

Close Modal

or Create an Account

Close Modal
Close Modal