Date:         Thu, 28 May 2009 09:26:24 -0500
Reply-To:     Warren Schlechte <Warren.Schlechte@TPWD.STATE.TX.US>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Warren Schlechte <Warren.Schlechte@TPWD.STATE.TX.US>
Subject:      Re: KS test in SAS
Content-Type: text/plain; charset="us-ascii"

I cannot vouch for one over the other.

However, please do note that using R, the Shapiro-Wilks test of
normality agrees with the SAS test for normality, with both stating
p=0.0179.

And, looking at a plot of the data suggests a very skewed dataset (12
points lie above the mean), even with only 18 points, so I would tend to
believe that the data are not normally-distributed.

Warren Schlechte

-----Original Message-----
From: Jinsong.Zhao@gmail.com [mailto:Jinsong.Zhao@GMAIL.COM]
Sent: Wednesday, May 27, 2009 10:40 AM
Subject: KS test in SAS

Hi there,

I using proc univariate to do KS test for a sample data set. However,
the p-value of KS test is different with that produced by SPSS. I hope
to know which one is correct.

Thanks in advance!

Regards,
Jinsong

The SAS code:

data x;
input x@@;
datalines;
3837 3334 2208 1745 2576 3208 3746 3523 3430 3480 3116 3428 2184 2383
3500 3866 3542 3278
;
proc univariate data = x normal;
var x;
run;
quit;

And the SAS results:

Kolmogorov-Smirnov    D     0.214278    Pr > D      0.0271

The SPSS code:

DATA LIST LIST /test .
BEGIN DATA
3837
3334
2208
1745
2576
3208
3746
3523
3430
3480
3116
3428
2184
2383
3500
3866
3542
3278
END DATA .

NPAR TESTS
/K-S(NORMAL) = test .

And the results is:

Most Extreme Differences        Absolute        .214
                                        Positive        .123
                                        Negative        -.214
Kolmogorov-Smirnov Z            .909
Asymp. Sig. (2-tailed)          .380