Extract from the Exploratory Data Analysis Code /* use proc freq to compute the number of levels for each question: task complexity */ PROC FREQ DATA = WORK.DATA12345C NLEVELS ; TABLES SQ01 - SQ59 / NOCOL ; ODS OUTPUT NLEVELS = WORK.QUESTLEVELS_1 ; RUN; /* sort (for future merge) and simplify the output data set produced by proc freq */ PROC SORT DATA = WORK.QUESTLEVELS_1 OUT = WORK.QUESTLEVELS_2 ( DROP = TableVarLabel ) EQUALS ; BY TABLEVAR ; /* question – task number */ RUN; /* explanatory variables: complexity of the task ( levels ) PER QUESTION distance function 1 and distance function2 */ /* target variable: DURATION PER QUESTION ( THE complexity of the task ) */ /* source data sets: QUESTLEVELS and DATA345DDSS */ /* PROC MEANS produces the ODS TABLE summary for USE IN the output data set */ /* exploratory data analysis for each of the three explanatory variables */ /* x1 = questlevels ; x2 = DL01 - DL59 ; x3 = DE01 - DE59: target variable = duration */ PROC MEANS DATA = WORK.DATA345DDSS STACKODS N MIN P25 MEDIAN MEAN P75 MAX MAXDEC = 2; VAR DQ01 - DQ59 ; /* compute median duration per question: sq01 to sq59 */ ODS OUTPUT SUMMARY = WORK.DATA3DP_STATS ; RUN; PROC MEANS DATA = WORK.DATA345DDSS STACKODS N MIN P25 MEDIAN MEAN P75 MAX MAXDEC = 2; VAR DL01 - DL59 ; /* compute median distance 1 per question: sq01 to sq59 */ ODS OUTPUT SUMMARY = WORK.DATADF1_STATS ; RUN; PROC MEANS DATA = WORK.DATA345DDSS STACKODS N MIN P25 MEDIAN MEAN P75 MAX MAXDEC = 2; VAR DE01 - DE59 ; /* compute median distance 2 per question: sq01 to sq59 */ ODS OUTPUT SUMMARY = WORK.DATADF2_STATS ; RUN; DATA WORK.DATA3DPM_STATS ( RENAME =( VARIABLE = QUESTION MEDIAN = DURATION ) ) ; SET WORK.DATA3DP_STATS ; RUN; DATA WORK.DATADF1M_STATS ( RENAME =( VARIABLE = QUESTION MEDIAN = DISTANCE1 ) ) ; SET WORK.DATADF1_STATS ; RUN; DATA WORK.DATADF2M_STATS ( RENAME =( VARIABLE = QUESTION MEDIAN = DISTANCE2 ) ) ; SET WORK.DATADF2_STATS ; RUN; DATA WORK.QUESTLEVELSM ( RENAME =( TABLEVAR = QUESTION NLEVELS_1 = NLEVELS ) ) ; SET WORK.QUESTLEVELS ; RUN; /* CHANGE THE VALUES OF THE QUESTION VARIABLE TO MATCH SQ01 TO SQ59 ALL 4 DATA SETS */ DATA DATA3DPQ_STATS ; /* No Quotes: DQ01-DQ59 TO SQ01-SQ59 */ SET DATA3DPM_STATS ; SUBSTR(QUESTION,1,2)="SQ"; RUN ; DATA DATADF1Q_STATS ; /* Single Quotes: DL01-DL59 TO SQ01-SQ59 */ SET DATADF1M_STATS ; SUBSTR(QUESTION,1,2)="SQ"; RUN; DATA DATADF2Q_STATS ; /* Double Quotes: DE01-DE59 TO SQ01-SQ59 */ SET DATADF2M_STATS ; SUBSTR(QUESTION,1,2)="SQ"; RUN ; PROC SORT DATA = WORK.DATA3DPQ_STATS OUT = WORK.DATA3DPS ( DROP = LABEL N MIN P25 MEAN P75 MAX )EQUALS ; BY QUESTION ; RUN; PROC SORT DATA = WORK.DATADF1Q_STATS OUT = WORK.DATADF1S ( DROP = LABEL N MIN P25 MEAN P75 MAX )EQUALS ; BY QUESTION ; RUN; PROC SORT DATA = WORK.DATADF2Q_STATS OUT = WORK.DATADF2S ( DROP = LABEL N MIN P25 MEAN P75 MAX )EQUALS ; BY QUESTION ; RUN; PROC SORT DATA = WORK.QUESTLEVELSM OUT = WORK.QUESTLEVELSS ( DROP = NLEVELS_2 NLEVELS_3 )EQUALS ; BY QUESTION ; RUN; DATA QUIZDATA ; MERGE QUESTLEVELSS DATA3DPS DATADF1S DATADF2S; BY QUESTION ; RUN; The final required data set looks like this : [image: Inline images 1] ##- Please type your reply above this line. Simple formatting, no attachments. -## ... View more

Hi,   For future reference, please mark the post of the person who correctly answered your question as the correct answer, not a reply to it that you have posted.  This helps keep acurate answers below the questions, and gives some credit to the poster who answered the question. Also noted on the post "Sum digits of a string". ... View more

Thank you very much KurtBremser and Kisharp. ... View more

Linus, sorry for not explaining my problem clear enough. However, I managed to get my desired data from code provided by both Kurt and Xia with only a slight change. Again thank you very much. ... View more

Thank you both so much! The solutions worked perfectly! ... View more

You are getting all "Germany" values except for . because of if ID in(.) then New_ID='N/A'; else New_ID='Germany'; ANY value of ID at this point other than . will yield "Germany". You want need:   if ID in(7:11) then New_ID='USA';   else if ID in(.) then New_ID='N/A'; else New_ID='Germany'; ... View more

Hi KC, Some answers: If the p-value of the interaction term is > 0.05 indicates the change for all levels of var1 is no difference. A non-significant interaction, but significant main effects, implies that the differences between the levels of var1 is the same at all times, and that the differences between levels of time is the same for all levels of var1.  Therefore, the best estimates of effects are the marginal means (main effect means). Does each of the significant simple difference tests by var1 reflect from the significant main time effect?   Yes, that is what is being shown. Also,  is it still sensible to quantify the simple differences and 95% C.I.  on the original scale as you previously posted here While you can report these, the marginal differences are probably more meaningful.  Try adding the following to your code: lsmeans time var1/diff cl ilink; to get the marginal means and differences. Steve Denham ... View more

Hi Steve. Yes, you are right. The estimates from both lsmeans and lsmeans are the same. My mistake due to oversight. Thank you very much. ... View more

Thank you all for your information ... View more