In PROC HPLOGISTIC, a data scientist specifies LINK=PROBIT in the MODEL statement instead of the default LINK=LOGIT. What change does this make to the model?
Select an answer to reveal the explanation.
Short Explanation
Here's the deal — when you choose the probit link, you get the cumulative normal distribution instead of the logistic sigmoid. Both keep predicted probabilities between 0 and 1, and they often give similar results.
Full Explanation
The cumulative normal distribution (probit) link uses the inverse cumulative standard normal distribution (Phi^-1) as the link function, modeling the probability as Phi(Xbeta) rather than the logistic sigmoid 1/(1+exp(-Xbeta)). Both produce probabilities in [0,1] and often give similar results in practice. The option saying it changes logistic regression to linear discriminant analysis is wrong because LINK= changes the link function, not the modeling algorithm. The option saying it forces probabilities outside [0,1] is wrong because both probit and logit links produce probabilities in [0,1]. The option saying it applies a log transformation to the target is wrong because LINK= affects the link function applied to the mean response, not a raw transformation of the target.