06-10-2014 09:21 AM
It seems as if you are fitting a discrete distribution to a normal distribution.
The results on my plot are good fit or not??
Impossible to say, without more context I would say this is a meaningless analysis
06-16-2014 03:57 PM
I'd first look at the plot of residuals vs the explanatory variables(s) to see if the residuals have any pattern (not good fit) or seem randomly distributed (good fit). That will be a better indicator of whether the fit is good.
The P-P plot is used to examine whther the residuals are normally distributed, which is not the same thing as a good fit. The doc for PROC UNIVARIATE describes how to interpret P-P plots: Base SAS(R) 9.4 Procedures Guide: Statistical Procedures, Second Edition
As Paige Miller indicates, it looks like you have repeated data or rounded data, so your residuals cannot be strictly normal, but they appear to be roughly normal, given that constraint.