glm for groups with unequal variance: is there an R equivalent of this Stata code? -
it well-known fact when fitting regression models groups, unequal variance between groups can cause trouble. there way in r account this?
a solution stata suggested in presentation http://www.stata.com/meeting/fnasug08/gutierrez.pdf (slides 9-17). actual code, shown on slide 14, follows:
# might specify . xtmixed . . . || id: boy agexboy girl agexgirl, nocons cov(un) # instead want . xtmixed . . . || id: boy agexboy, nocons cov(un) || id: girl agexgirl, nocons cov(un)
is there similar solution in r?
Comments
Post a Comment