@Ruhi wrote: I resolved my problem by adding quotes around like this, CR= "&CR"; But I would like to understand why this happened. If anyone can help.   Thanks. Look at the symbolgen messages for ID, LDMK and SIDE. They all contain the quotes tha are needed where they are used. CR does not contain quotes. Since having quotes in macro variables often causes problems, I would define all those variables without quotes and wrap them in double quotes where needed. ... View more

Why not use BY groups in your calculation data step so that you can calculate DISP, FORCE, ACC, VELOCITY etc in one data step for all of the groups of data?   Instead of using (_N_=1) in your logic you could use a FIRST. variable to test when a new group starts? ... View more

Your description does not give us enough information to determine whether the statistical model is correct. For example, how many levels of CASE are there, and how does CASE relate to TRIAL? Are there 3 CASEs with one TRIAL each? Or 3 TRIALs for each CASE? What research question do you have that would be addressed by 12 SCANs in each TRIAL?    Your residual plot and data plot show that there is an upper bound (which is zero) to your response "PlumbResult_0504_LateralDeviatio". Neither the lognormal nor the gamma distribution is appropriate for data with an upper bound; both the lognormal and the gamma have a lower bound at zero and an upper bound of infinity. Both should have failed miserably with a response with negative values (the log of zero is not defined, and I would guess that there was a message to that effect in the log window; always pay attention to the log window).   So, we need to know more about what your response is measuring, in addition to more about your experimental design. Guessing wildly, you might have more luck redefining your response as (-1)*response, if that was sensible in context; that redefined response might follow the exponential distribution, and then the gamma might work (although gamma mixed models can be very persnickety).    Given your current description, I doubt your RANDOM specifications are right but we await more information.   ... View more

Here is the distribution after subtracting the means from each condition. It is still multi-modal.  ... View more

Why not introduce the subgrouping as a fixed effect, and include it in a group= option to the random statement.  For example, if sex were the factor, you might try:   proc mixed data=required covtest cl; class pt day scan sex; model var= sex; random pt day(pt) scan(day*pt)/group=sex ; estimate "Mean" intercept 1/cl; ods output covparms=Covparms estimates=EstimatedMean; run; Steve Denham ... View more

Why do you need a macro for that?  You will be aware of by group processing, set your data up correctly and you could simply by group the whole dataset: proc mixed data=prep method=type1 cl; by <group variables>; class pt image scanwithinday ; model <variable>=; random pt scanwithinday(pt); run; Far simpler and more efficient.   ... View more

Since Excel is CELL focused and SAS is Column oriented then one of the things that sometimes makes Excel to SAS work better is to ensure that the same cell format is applied to all cells in a column. One of the obnoxious things with Excel is that somethings as entered become a numeric value with an Excel date or time format while others are character values. Try this for example: 1) in a new Excel spreadsheet type:  1/1/2016 then move to another cell. 2) return to that Cell and look at the format (Right click, select format cells, go to the number tab). It will be a DATE type with a date format on the right you could pick. 3) go to a different cell in same column and type '1/1/2016  (note the apostrophe) 4) do the same examination, You will see a "general" format. Interesting results when importing that sheet into SAS will result depending on approach taken.   One thing you can try: highlight the entire column in your sheet and set the format to NUMBER (not date). Any character values will be noticeable due to dashes or /. Retype those as 1/1/2015 without the apostrophe any will see those appear as a number (due to the previous setting for the column). After you have addressed any of those changes needed in your data then reset the entire column format to date. If you have multiple date fields it is recommended to set them all to the same appearance. Then import the resultant file. (or save to CSV and import that one). Of course if you have lots of records this may not be practical in which case more creativity is needed. But you really do need to look out for data that is mixing  mm/dd/yy and dd/mm/yy formats in the same column. ... View more

Which type of correlation were you running? Response to Pre "question A" with Post "question A" or  pre/post difference question A with pre/post difference of question B?   If you look at the pre and post values for the same questions correlation may not be very helpful. Here is an example why: data junk; Do Pre = 1 to 3; do mult = 1 to 3; do post = (pre+1) to 5; output; end; end; end; drop mult; run; proc corr data=junk;run; proc freq data=junk; tables pre*post/list nocum nopercent; run; This is an artificial data set. In every case the Post value is greater than the Pre value. The Pearson correlation is about .4 but the data   is made to show 100 percent improvement or decline (depending on interpretation) because the Intervals 1 to 2 and 1 to 5 differ. ... View more

I still don't understand your question. Sorry. The SE is the measure of variability (uncertainty) for the estimated parameter of a covariable. It is printed right next to the parameter estimate in the solution table (just use the  "/ s" option on the model statement. You can use to get confidence intervals. One does not multiply the SE by anything related to df. ... View more

Hi Ruhi, I think you have a confounding problem, based on your description.  It appears from the contingency table that the variable "direction" is completely confounded with landmark (i.e., three of the landmark descriptions look like anterior and the other three are posterior).  However, since you are already fitting a means model, "direction" can be extracted later from the lsmestimate statement--once we get convergence.  So I would consider:   proc glimmix data=dat1 noitprint method=laplace;   class absasym  landmark  ;   model Autoscore(desc) = landmark*absasym / dist=binary link=logit;   random intercept / subject=ID gcorr solution;   nloptions technique=congra maxiter=1000 gconv=1e-4; run;   See if this will converge.   Steve Denham ... View more

If you really merge by id AND ldmk the you'll get the expected result - at least with the sample data you've posted. Run the code below and you'll see. data peakpressure;   infile datalines truncover dlm=' ';   input (ID ldmk SIDE1) ($);   datalines; AE50 L1 RIGHT AE50 L1 LEFT AE50 L2 RIGHT AE50 L2 LEFT AE50 L3 RIGHT AE50 L3 LEFT AE50 L4 RIGHT AE50 L4 LEFT AE50 L5 RIGHT AE50 L5 LEFT ; run; data efinding;   infile datalines truncover dlm=' ';   input id $ ldmk $ respMT :best32.;   datalines; AE50 L1 -1 AE50 L2 1 AE50 L3 1 AE50 L4 1 AE50 L5 -1 ; run; proc sort data=peakpressure out=peakpressure2;   by id ldmk; run; data merged_respmt;   merge peakpressure2 efinding;   by id ldmk; run; ... View more