The execution time will depend on which tests are chosen.  A few, in fact, have explicit limits on sample size allowed, but those would issue an error message and stop.

I ran a test with just Anderson-Darling, which is the default univariate, and 26600 cases, and it completed in one minute.  Then I ran a test with all the allowed univariate (excluding Shapiro-Wilk) and several multivariate tests, and that took much longer.  The test tables appeared in about 4.5 minutes, but there was a problem with the plots (boxplot, histograms, qq and splom).  I will investigate that.

You should be using the updated version of the command, which is what you would have if you installed it recently, as the original version was broken by changes made by the author of the original underlying R module.

I have discovered that the Shapiro-Francia test has an undocumented limit of 5000 cases.  I have added a guard on that to the NORMALITY ANALYSIS procedure.  If I remove that from the univariate tests list, everything works and completes in 4.5 minutes.

I am using Stats V. 31.0.

Sending files to your Gmail.