Re: [Gretl-users] logit model - f(beta'x) at mean of independent vars

Am 21.06.2011 15:24, schrieb Katrin Woelfel: > Dear Gretl community, > > When working with a logit model (binary) in GRETL, I wondered about the > number of "f(beta'x) at mean of independent vars" in the output. > Can anybody tell me, how this value is computed? > Thanks a lot in advance. > Hi, well the x are the indep vars. The beta are the coefficients for which you have estimates. So it very much seems to be f(\hat(beta)'\bar(x)). The final piece is f(), and here's my quiz question for you: if you do logit, which function is used for f()? ;-) cheers, sven