Hi

I ran a logistic regress in SAS but did not quite understand the meaning of Residual Chi-square Test. Could anybody help I would appreciate.

Are you talking about The Hosmer-Lemeshow Goodness-of-Fit Test?

Are you talking about The Hosmer-Lemeshow Goodness-of-Fit Test?
Yeah. I found yesterday that the bigger the better, as long as it =>0.05. Yes?

Hosmer-Lemeshow test is a (summary level) goodness of fit test. Not a residual test.

Here the assumption is the model is perfect.
So if the model is perfect then the chisquare value ( obs value - actual value =0) become zero. and P value become 1.
So higher the P value ( smaller the chisquare value) is better.

Thank you friend, it is much clearer now :)