I need someone without EXACT TESTS run some syntax

[Bruce Weaver]

Good morning folks.  I would appreciate it if someone who does not have the EXACT TESTS module would run the following code, and report back on the results they get.

DATA LIST free / Group SOB (2F2.0).
BEGIN DATA
1 8   1 5   1 7  1 4  1 12
2 3   2 6   2 0  2 2  2 1
END DATA.

VALUE LABELS Group 1 "Placebo" 2 "New Drug" .
VARIABLE LABELS SOB "Shortness of breath: # of episodes".

* Include /METHOD=EXACT.
NPAR TESTS
  /M-W= SOB BY Group(1 2)
  /MISSING ANALYSIS
  /METHOD=EXACT TIMER(5).

* Exclude /METHOD=EXACT.
NPAR TESTS
  /M-W= SOB BY Group(1 2)
  /MISSING ANALYSIS
.

I do have EXACT TESTS.  When I run this code, I get the following output:

Test Statisticsa
        Shortness of breath: # of episodes
Mann-Whitney U  2.000
Wilcoxon W 17.000
Z -2.193
Asymp. Sig. (2-tailed) .028
Exact Sig. [2*(1-tailed Sig.)] .032b
Exact Sig. (2-tailed) .032
Exact Sig. (1-tailed) .016
Point Probability .008
a Grouping Variable: Group
b Not corrected for ties.

Test Statisticsa
        Shortness of breath: # of episodes
Mann-Whitney U  2.000
Wilcoxon W 17.000
Z -2.193
Asymp. Sig. (2-tailed) .028
Exact Sig. [2*(1-tailed Sig.)] .032b
a Grouping Variable: Group
b Not corrected for ties.

Notice that the second lot of output is the same as the first except that the last two Exact Sig. lines and the Point Probability line are omitted.  

Some students in my class who do NOT have EXACT TESTS tell me they are getting different results.  One of them sent this output:

Test Statisticsa
        Shortness of breath: # of episodes
Mann-Whitney U  3.000
Wilcoxon W 18.000
Z -1.997
Asymp. Sig. (2-tailed) .046
Exact Sig. [2*(1-tailed Sig.)] .056b
a Grouping Variable: Group
b Not corrected for ties.

Are others who do not have EXACT TESTS getting this result too?  If so, can anyone offer an explanation?  

Thanks,
Bruce

[Andy W]

At work I have V22 (No exact tests), and I get the same results you are reporting Bruce.

[Bruce Weaver]

Thanks Andy.  

Andy W wrote

At work I have V22 (No exact tests), and I get the same results you are reporting Bruce.