I have a data set (final) that I am trying to predict values from. I would like to predict writing score (write) from each permutation of math=70,90 and social studies (socst)=70,90. My thoughts were to create a new data set with the desired math and social studies combos, concatenate the two, and then run proc reg with 95% CI (clb) to get the predicted values. However, while my combined data set has these values on my proc print, it has blanks for predicted values. Any idea what I'm doing wrong? Thank you! *create new data set; data new; do math=70,90;    socst=70;   output; end; do math=70,90;    socst=90;    output; end; run; *join data sets' data final_pred1; set new final; run; *use proc reg to get predicted writing scores; proc reg data=final_pred1; model write = math socst/clb; run; quit; proc print data=final_pred1; run; ... View more

I am trying to merge two data sets and then create new variables from these data sets. In my library, I have two data sets (cross1 and long1) with the common variable MEDRNO. I'm trying to merge them and create two new variables (change_dilation and change_time). For these new variables, i want to take the difference between the first observation of dilation and all other observations of dilation. The same goes with change_time. However, I'm not getting any observations in my output. Any idea what I am doing wrong? *question 1; libname Hwk4 'C:\Users\USERABC\Desktop\Assignment 4'; run; *question 2a; proc sort data=hwk4.cross1;by MEDRNO; proc sort data=hwk4.long1;by MEDRNO; data merged; merge hwk4.cross1 hwk4.long1; by MEDRNO; run; *question 2b; data hwk4; if first.MEDRNO = 1 then do;             base_dil = DILATION;             t0 = time;       end; retain base_dil t0; change_dilation = DILATION - base_dil; change_time = time - t0; run; proc print; run; ... View more