the effect of taking AP Economics on performance on the economics EOCT for those who take an AP Economics course).
We observe the first term on the right-hand side of Equation 1, the economics EOCT scores for students who actually took AP Economics.
Once we generate the matches between treatment and control students, we then calculate the average economics EOCT score for these control students (D = 0) who have been matched to our treatment students (D = 1) and use this average as an estimate of the second term on the right hand side of Equation 1.
Because we have a strong measure of students' prior achievement in the form of EOCT scores in algebra and geometry, we believe that ITA holds.
Below, we provide evidence that controlling for unobserved school quality does not seem to impact the estimated effects of AP course taking on Economics EOCT scores--once we control for unobserved teacher quality.
To create baseline estimates and to establish the validity of our matching variables, we first estimate a linear regression (Equation 2) that models each student's standardized economics EOCT score ([EOCTScore.
t], a vector of shift parameters that captures the effect of time-invariant teacher characteristics on student economics EOCT scores:
The OLS results imply that students who choose to take an AP class in economics score much better on the economics EOCT than do students who do not take an AP class in economics.
estimates of the effect of AP enrollment on EOCT scores.
Given that the estimated AP effects increase when we include only school fixed effects, we infer that unobserved school quality does not seem to bias upwards estimates of AP effects--controlling for prior achievement in mathematics, race and class, and the other variables used to explain Economics EOCT scores.
The estimates of the effects of the control variables on Economics EOCT scores reported in Table 2 have the expected signs.
We also match students based on algebra EOCT score and geometry EOCT score because scores on these previous tests should be reliable proxies for ability and work ethic--and prior research has found that they are highly correlated with performance in economics.