Xavier Viviand asked by Stata did not reproduce the results published in a textbook:
"I am trying to replicate with Stata the two way Between-Within Analysis of
Variance example proposed by Keppel (Design and Analysis, 3rd ed, Table 17-6
pp 375)."
If it's any help in identifying the source of the problem, when I input the data from
Xavier's posting, I get exactly the results in Keppel's textbook. (I've copied the data
from Xavier's posting, pasted them into a text file, insheeted them into Stata and
issued two -reshape, long- commands before copying the -list, noobs- output from
Stata's Results window and pasting the dataset into the do-file.)
Joseph Coveney
-----------begin keppel.do-------------
set more off
clear
input byte a byte s byte b byte y
1 1 1 13
1 1 2 14
1 1 3 17
1 1 4 20
1 2 1 10
1 2 2 11
1 2 3 15
1 2 4 14
1 3 1 13
1 3 2 19
1 3 3 18
1 3 4 21
1 4 1 4
1 4 2 12
1 4 3 14
1 4 4 15
2 1 1 5
2 1 2 10
2 1 3 17
2 1 4 21
2 2 1 8
2 2 2 15
2 2 3 22
2 2 4 24
2 3 1 14
2 3 2 16
2 3 3 23
2 3 4 23
2 4 1 12
2 4 2 21
2 4 3 26
2 4 4 26
3 1 1 13
3 1 2 24
3 1 3 28
3 1 4 30
3 2 1 9
3 2 2 22
3 2 3 22
3 2 4 24
3 3 1 14
3 3 2 22
3 3 3 28
3 3 4 28
3 4 1 8
3 4 2 18
3 4 3 27
3 4 4 28
end
anova y a / s|a b a*b, repeated(b)
exit
-----------end keppel.do-------------
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
