Correct, I do need the trimmed data per BY group in a data set with the outliers either deleted or set to missing. Within the output for proc univariate or proc mean can you specify an output argument similar to the below that would work? Where I want the descriptive statistics for all the variables within my array (x_1, x_2, x_3.... x_56) but I want a seperate table for each variable. proc mean data= ; by age; var x_1 x_2 x_3... x_56; output IF var = x_1 THEN output=X_1 mean=average n=n lclm =lower uclm=upper ; IF var = x_2 THEN output=X_2 mean=average n=n lclm =lower uclm=upper ; IF var = x_3 THEN output=X_3 mean=average n=n lclm =lower uclm=upper ; .... IF var = x_56 THEN output=X_56 mean=average n=n lclm =lower uclm=upper ; run;
... View more