I am trying to find a way to calculate the McFadden R2 for a tobit regression, but I cannot find the right package/formula. I now tried it via the following:
fm0 <- update(fmtobit2, . ~ 1)
1 - as.vector(logLik(fmtobit2)/logLik(fm0))
But this gives me a value above 1, does anyone know another method of how I could calculate this value?
Thank you in advance!