Hello,
My question is about areg, or hierarchical regression. STATA says
hierarchical regression (areg) is "regression with a large dummy variable
set". It is not my understanding that hierarchical regression must involve
dummy variables; can anyone provide any explanation of this? Thanks very
much,
Erin
Where'd you get the hierarchical part? areg is described as "Linear
regression with a large dummy-variable set." So if, say, you've got a
categorical variable with 40 categories, it will be "absorbed", i.e. the
results will be right but you won't see the coefficients for each dummy
variable. Besides keeping your output from being cluttered, it may let you
run bigger regressions than you would be able to otherwise.
