Is there any way to rescale the constant in an OLS regression? I had to
rescale one independent variable because the differences in the units of
measurements caused huge standard errors in that variable. The rescaling
solved the problem for that particular variable, but I have still the same
problem of a very large coefficient and standard error for the intercept
term. Is there any way to deal with this (what I would basically like to do
is to multiply the constant by 1000)?
