search

konfound-project / konfound

R package for carrying out sensitivity analysis
https://konfound-project.github.io/konfound/
Other
15 stars 7 forks source link

Model support #49

Open ben3mey opened 11 months ago

ben3mey commented 11 months ago

Currently, the package only supports lm() models. Will there be an update to support further regression models such as plm() for example?

kenfrank1 commented 11 months ago

appreciate the interest. We'd need to know more about the specific model you want to estimate (e.g., use of fixed or random effects, centering, natural log). as a starter, you can check out the FAQ which may address some of your issues via a multilevel model.

ben3mey commented 11 months ago

Thank you for your answer! Specifically, I would like to use your package for my plm() models with fixed effects. Here is the exact model specification calculated in R:

m1 <- plm(dv ~ iv +
            controls +
            factor(year),
          data = df,
          index = c("gvkey", "year"),
          model = "within")

dv & iv = continuous and numeric variables gvkey = firm identifier df = unbalanced panel data set

Hope that helps! I saw in the FAQ already that this is under development. Appreciate the effort. Cheers!

kenfrank1 commented 10 months ago

The FAQ is the best we can do for now.