0

I am trying to estimate Treatment Effects and robust standard errors after propensity score matching with two continuous outcomes following this guide: https://cran.r-project.org/web/packages/MatchIt/vignettes/estimating-effects.html

I performed nearest neighbor pair matching without replacement (MatchIt package). I used the code below for my data, However it does not seem to work with two dependent variables. This is the error message I'm receiving: Error in SSD.mlm(object) : Weighted 'mlm' objects are not supported

I am grateful for any tips and suggestions on alternative procedures

install.packages("sandwich")
library(lmertest)
library(sandwich)

fitm <- lm(cbind(PHQ_9_Sum,SF8_Sum) ~ parenthood, data = mdata, weights = weights)

coeftest(fitm, vcov. = vcovCL, cluster = ~subclass)
Janni
  • 1
  • 1
  • Hi Janni and welcome to stack overflow. Questions are more likely to get answers if they contain a [minimal, reproducible example](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example) - this with R requires not only the code, but also the data required to reproduce the error. Please edit your question to include your data in code form - see link above if you need help on how to do that. – Andrea M Aug 03 '22 at 18:23
  • Janni, why can't you just fit to separate outcome models, one for each outcome? Why does it have to be a multivariate model? – Noah Aug 03 '22 at 18:37

0 Answers0