- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Dear all, From F table, could find the F(7,10),0.05=3.14, but SAS output (see the bottom of picture) Pr> F(=3.65) = 0.0639. There was something wrong ?* SAS code;data unequal_var;
input disease $ time @@;
cards;
A 36 A 32 A 24 A 35 A 26 A 28 A 32 A 33 A 34 A 29 A 32
B 35 B 24 B 13 B 27 B 22 B 18 B 31 B 30
;
proc ttest;
class disease;
run;
Accepted Solutions
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Hello @chchao and welcome to the SAS Support Communities!
Everything is correct here (as usual). The keyword is "Folded F" -- which means that a factor 2 comes into play in order to perform a two-tailed test. The formula can be found in section "Two-Independent-Sample Design" of the PROC TTEST documentation. To reproduce the p-value (0.0639) with SAS functions you can use PROBF or CDF¹: p=2*(1-probf(230/63, 7, 10))=2*(1-cdf('F', 230/63, 7, 10)). (230/63 is the exact F value 3.65079...)
¹ Addendum: ... or SDF: 2*(sdf('F', 230/63, 7, 10))
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Hello @chchao and welcome to the SAS Support Communities!
Everything is correct here (as usual). The keyword is "Folded F" -- which means that a factor 2 comes into play in order to perform a two-tailed test. The formula can be found in section "Two-Independent-Sample Design" of the PROC TTEST documentation. To reproduce the p-value (0.0639) with SAS functions you can use PROBF or CDF¹: p=2*(1-probf(230/63, 7, 10))=2*(1-cdf('F', 230/63, 7, 10)). (230/63 is the exact F value 3.65079...)
¹ Addendum: ... or SDF: 2*(sdf('F', 230/63, 7, 10))