Hi, I'm doing a difference-in-differences analysis of a health policy intervention.
Health clinics are paid for the percentage of eligible patients who they give the right treatment to. The clinics are measured/paid separately for each type of treatment indicator.
One year, some of the indicators stopped being paid (but others continued). I want to measure if stopping the payment had an impact on performance (relative to the control group).
I have data for every single health clinic in the country - for the intervention and control group (both pre- and after- intervention.
eg.
Intervention 2016 Intervention 2017 Control 2016 Control 2017
Clinic1 82.20% 86.14% 78.90% 78.79%
Clinic2 70.70% 68.23% 78.24% 72.46%
Clinic3 67.74% 54.22% 72.28% 64.95%
Clinic4 82.26% 87.87% 91.18% 85.39%
… … … … …
up to Clinic 450
Average 78.17% 74.14% 78.30% 75.98%
So the difference-in-differences estimator (the intervention impact) is: ((Intervention2017-Control2017)-(Intervention2016-Control2016)) = -1.97 (based on the average)
My questions are:
1. If I have data for every single health clinic (it's not a sample), should I still calculate the Standard Error and confidence intervals for the DiD estimator?
2. How best should I calculate the SE and CI for the DiD estimator? Should I use just the mean average, or can I factor in the individual pairs of measurements (for each clinic)?
3. Each clinic has varying numbers of eligible patients (I have the numerator/denominator for each); can I factor that into the estimation?
4. Can I also calculate whether there's a statistically significant change in performance variation between practices, following removal of the payment incentive? How best to calculate that?
Ideally I'd like the do the calculations on STATA.
Many thanks.
Health clinics are paid for the percentage of eligible patients who they give the right treatment to. The clinics are measured/paid separately for each type of treatment indicator.
One year, some of the indicators stopped being paid (but others continued). I want to measure if stopping the payment had an impact on performance (relative to the control group).
I have data for every single health clinic in the country - for the intervention and control group (both pre- and after- intervention.
eg.
Intervention 2016 Intervention 2017 Control 2016 Control 2017
Clinic1 82.20% 86.14% 78.90% 78.79%
Clinic2 70.70% 68.23% 78.24% 72.46%
Clinic3 67.74% 54.22% 72.28% 64.95%
Clinic4 82.26% 87.87% 91.18% 85.39%
… … … … …
up to Clinic 450
Average 78.17% 74.14% 78.30% 75.98%
So the difference-in-differences estimator (the intervention impact) is: ((Intervention2017-Control2017)-(Intervention2016-Control2016)) = -1.97 (based on the average)
My questions are:
1. If I have data for every single health clinic (it's not a sample), should I still calculate the Standard Error and confidence intervals for the DiD estimator?
2. How best should I calculate the SE and CI for the DiD estimator? Should I use just the mean average, or can I factor in the individual pairs of measurements (for each clinic)?
3. Each clinic has varying numbers of eligible patients (I have the numerator/denominator for each); can I factor that into the estimation?
4. Can I also calculate whether there's a statistically significant change in performance variation between practices, following removal of the payment incentive? How best to calculate that?
Ideally I'd like the do the calculations on STATA.
Many thanks.