Assuming that you want an estimate (point and/or confidence interval) of the difference in those two probabilities, then yes, that code is reasonable (though you don't need the ILINK option when the identity link is used).  If you use the default logit link then the differences from the LSMEANS statement will be differences in log odds (equivalently, log odds ratios).  See the following note for more, in general, on estimating differences in probabilities:    37228 - Estimating differences in probabilities with confidence interval ... View more

Hello K C, If you have SAS/STAT13.2, you can get Bowker's test output (generalization of McNemar's test to 2xK tables) with: proc freq data=tmp; tables before_level10*after_level10 / agree; run; I can't guarantee this is what you are looking for, but it is now available. Steve Denham ... View more

When i checked his paper i did not find any reference to the code " A Rake-Trim SAS® Macro and Its Uses at Westat ", you may want to ask him. You can find his mail at the end of the paper. Although i think it is in the property of Westat ... View more

Dear Keshan and Steve, Thanks for your reply. I wish to analyse only the 2013 data and to create weighting for the 2013 only participants so it will be a nationally representative of the population. However, I'm not sure how do I go about doing the weighting given that there are only drinkers in the 2013 data. Thank you. ... View more

You can also now compare two SAS programs within SAS Enterprise Guide -- new in v7.1: How to compare SAS programs in SAS Enterprise Guide Chris ... View more

Thanks for your reply. I checked that there is neither a missing column nor invalid SAS naming conventions. I managed to read in all my data by proc import with GUESSINGROWS=2147483647; statement and I still don't understand why I can't do it with libname statement :smileyconfused: Because I need to meet deadline I'll just use proc import for now ... View more

I agree, especially if the denominator is large.  However, if the denominator is less than 50, then the genmod code would be my preference. Steve Denham ... View more

  timegrp= hour(time/2) +1; ... View more

Could use PROC FREQ and the CHISQ? Chi-sq will tell you that at least two of the proportions differ. Anca. PS:you may need to re-arrange your data something like: data have; input sample $ ethnicity $; cards; A white A white B black B white B asian A asian A asian B black ; proc freq data = have; tables sample*ethnicity/chisq; run; Anca Message was edited by: Anca Tilea And this link actually offers more details and probably what you need: http://support.sas.com/kb/22/565.html ... View more

Thanks a lot everyone for your help! Great to know there are many different ways to achieve this. ... View more

How about this: proc sql;     select *,sum(flag ne 0) as final2     from a     group by id; quit; Good luck! Anca. PS: and if you want a table, then add the proc sql;      create table your_table as     select *,sum(flag ne 0) as final2     from a     group by id; quit; ... View more

thanks guys ... View more

That model ought to be interpretable. Steve Denham ... View more

Thanks Steve for your reply! Much appreciated! I'll give it a go as you suggested. ... View more