Closed darrgerr closed 7 months ago
At this point, I got the answer. I should have added the part 'lead = 0:4' to get the estimates for all lags.
PM.results <- PanelMatch(lag = 4, time.id = "year", unit.id = "new_id",
treatment = "appointed_new", refinement.method = "mahalanobis",
data = dtb_psm, match.missing = TRUE,
covs.formula = ~ vpm + n_staff_mon,
verbose = TRUE, exact.match.variables = list("special_status", "u_spec"),
size.match = 5, qoi = "att" ,outcome.var = "growth_f",
lead = 0:4, forbid.treatment.reversal = FALSE, use.diagonal.variance.matrix = TRUE)
PE.results <- PanelEstimate(sets = PM.results, data = dtb_psm, se.method = "bootstrap", number.iterations = 1000, confidence.level= .95)
PE.results[["estimates"]]
I'd like to thank the authors of the package for its introduction. I also wish to clarify the issue I've been dealing with within the last few days.
I am trying to estimate the effects of appointment of a university leader (appointed_new) on university's performance (growth_f), I run the following code:
When I run covariate balance (I have also done it with no refinement method), I can see output for all 4 lags:
However, when I use PanelEstimate, the output is rather strange:
I have searched for potential reason everywhere, rerun my code looking for a typo, however, could not find an issue. Hence, I thought that the problem might be within the package itself. Is it the case? Ready to provide code and data. Thank you