other than proc ttest is there a way to apply two sample two sided t-test

[chinaski]

hi,

my enterprise guide do not have sas/stat licence so proc ttest is not working, is there another procedure for two sample t-test. You can give the codes from testscores data. Thanks.

By the way it is very annoying that one programme has many packages, we are trying to learn a programme here, everytime another obstacle comes along.

[FreelanceReinh]

Hi @chinaski,

SAS University Edition includes SAS/STAT among other modules, so this would be a great option for learning purposes.

Otherwise, the two-sample t-test is simple enough to implement the formulas found in the documentation in SAS/Base code.

Example including mean difference, pooled standard error, t-statistic and two-sided p-value:

proc summary data=sashelp.class nway;
class sex;
var height;
output out=stats n=n mean=m std=s;
run;

data want(keep=d se t p);
set stats;
if _n_=1 then do; m1+m; n1+n; s1+s; end;
else do;
  d=m1-m;
  se=sqrt(((n1-1)*s1**2+(n-1)*s**2)/(n1+n-2)*(1/n1+1/n));
  t=d/se;
  p=1-probf(t**2,1,n1+n-2);
  output;
end;
format p pvalue.;
run;

Result:

    d          se          t         p

-3.32111    2.28628    -1.45263    0.1645

For comparison:

proc ttest data=sashelp.class;
class sex;
var height;
run;

Partial output (relevant numbers highlighted):

The TTEST Procedure

Variable:  Height

Sex           Method             N        Mean     Std Dev     Std Err     Minimum     Maximum

F                                9     60.5889      5.0183      1.6728     51.3000     66.5000
M                               10     63.9100      4.9379      1.5615     57.3000     72.0000
Diff (1-2)    Pooled                   -3.3211      4.9759      2.2863
Diff (1-2)    Satterthwaite            -3.3211                  2.2883

Sex           Method               Mean       95% CL Mean        Std Dev      95% CL Std Dev

F                               60.5889     56.7315  64.4463      5.0183      3.3897   9.6140
M                               63.9100     60.3776  67.4424      4.9379      3.3965   9.0147
Diff (1-2)    Pooled            -3.3211     -8.1447   1.5025      4.9759      3.7339   7.4596
Diff (1-2)    Satterthwaite     -3.3211     -8.1551   1.5129

Method           Variances        DF    t Value    Pr > |t|

Pooled           Equal            17      -1.45      0.1645