then if I use proc ttest, I do not need to do proc mixed test?   ... View more

I am testing 750 data points   the column are in year, age sex   the rows are group of year 1900-1910...ect it goes all the way up to 1990-2000 the data is mortality rate   I wanted to test the mean between female and male  from yer 1900 till 2000 I wanted to know if mortality for female are higher than male,   should I use  proc glm or  proc ttest. in the example below   proc glm data = Project; class sex; where sex in ("f" "m"); model rate = sex; run; quit;   or proc ttest data = Project; class sex; var rate; where sex in ("f" "m"); run; quit;       then 2nd test there are 10 years grouping and each years there are 100 mortality rate points. I would like to do a pairing comparison for each 10 years. to determine if there was a significant changes. I am not sure about the code . it could look like that I should have 10 groups and tukey test will tell me if there is a significant between years group   I cannot set it up correctly. I attached my sas code   proc glm data=project; class sex; where sex in ("t"); model rate = year sex / solution; means year sex/ tukey; run; quit; ok see if you can help         ... View more

sorry for the confusion I want to do a t test using multiple class for the column and row do I use proc glm class column, row var data     is it right? ... View more

sorry the code I used is this proc means data=project mean ; class sex age ; var rate  ; where age in (4 5) & sex in ("t" ); run; proc ttest data = Project; class sex age; var rate ; where  age in (4 5) & sex in ("t"); run; quit;   the ttest did not work  prob with "sex" in class run; ... View more

compute mortality rate for female and male for various year groups   I have 9 years group in column 1 female column 2  rate column 3  male column 4 and rate column 5   the class is  on year  female rate male rate var has rate I need to add male and female   help? ... View more