I am using
HTML Code:
xtreg
to run an event study model and would like to use coefplot with it. I have used regress command to generate plots of coefficients using coefplot with no issues. However, now I have the following regression:
HTML Code:
xtreg outcome ib9.time_indicator##ib0.treatment [w = weight], fe robust cluster(state)
, allowing me to use analytic weights to ensure estimates reflect the sample. When using this specification with coefplot (code:
HTML Code:
coefplot drop(`vars') vertical yline(0)
), it is clear that the coefficient estimates visualized with coefplot don't match the values in the weighted regressions. Given the propietary nature of my data, I can't share an extract, but wanted to ask: have folks used weights with coefplot effectively so that estimates from the weighted regression and coefplot align?