Can someone provide an example of how to calculate power/sample size for cochran mantel haenszel? The response variable is binary and there are two stratum (one has 3 levels and the other is binary). ... View more

I'm trying to use MIANALYZE on the ODS output dataset called ROCASSOCIATION that comes out of the logistic procedure.  I'm trying to combine multiple AUCs.  Help please!                   ods output ROCAssociation=ROCAssociation;                   proc logistic data=both plots=roc(id=prob);                         by _imputation_;                      model y(event='Yes') = x / nofit ;                      roc 'X' x;                   run; Then... PROC MIANALYZE ?????? ... View more

I'm using the code below and want DAYS to be summarized in order from day -28 to day +28, however for the negative days, each page orders the days like they were positive.  For example page one of the table goes from day -21 to day -28 instead of from day -28 to day -21, etc.   proc report data=final out=tvalid.v_&pgmname nowd headline split='|' missing/* style = [outputwidth=100%] */ style(header)=[asis=on just=center protectspecialchars=off] style(column)=[asis=on just=center protectspecialchars=off] ; column pgbrkvar period periodc diary_day order_ stat_cat col1; define pgbrkvar / order noprint; define period / order noprint; define periodc / order style=[just=l cellwidth=1.1in] "Period"; define diary_day / order style=[just=l cellwidth=1.1in] "Day"; define order_ / order noprint; define stat_cat / order style=[just=l cellwidth=1.1in] "Statistic"; define col1 / display style=[just=c cellwidth=0.7in] "(N=&g.)"; break after pgbrkvar /page; compute before; line ' '; endcomp; run; ... View more

how can I get the 95% confidence interval for the difference between two medians? ... View more

SAS v9.4. Can I use proc power - onesamplefreq to calculate sample size for a single proportion performance goal hypothesis? I want to show success rate is at least 81%.  I expect the sample to show 90% success and want sample size to ensure the lower limit of the one-sided 95% confidence interval is at least 81% (alpha=5%, power=80%).  Help!  Is this correct?   proc power ; onesamplefreq  sides=1 alpha=0.05 nullproportion=0.81 /*Performance Goal-PG*/ proportion= 0.90  /*must be slightly higher then PG*/ ntotal= . power= .8 ; run; ... View more

How would you conduct propensity score analysis for an observational set of data (two non-randomized treatment groups and multiple baseline characteristics), but we want to be blinded to the outcome.  Would we just stop after calculation of the propensity scores?   Thank you for your insight! ... View more

Does anyone know the SAS code for non-inferiority test for VARIANCES of two group?  I see lots on means and proportions but can't find help on variances.  Thx! ... View more

Want to show adverse event outcome is no greater than 10%.   1) What is the code for sample size/power calculation?   2) What would the hypotheses be?   3) Can we present it in confidence intervals? ... View more