 |
This web page was last updated at 11/16/99 04:11 PM
-
Suppose there are 6 different treatments with the
following sample means:
|
|
A
|
B
|
C
|
D
|
E
|
F
|
|
Mean
|
21
|
12
|
15
|
17
|
19
|
10
|
The Bonferroni multiple comparison procedure
conducted 15 hypothesis tests with the following results (D=Don’t
Reject Ho and R=Reject Ho):
|
|
A
|
B
|
C
|
D
|
E
|
F
|
|
A
|
|
R
|
|
R
|
D
|
R
|
|
B
|
|
|
D
|
D
|
D
|
D
|
|
C
|
|
|
|
D
|
D
|
R
|
|
D
|
|
|
|
|
D
|
R
|
|
E
|
|
|
|
|
|
R
|
|
F
|
|
|
|
|
|
|
Order the treatments by their means (from smallest to largest) and
draw bars over treatments that are not significantly different.
-
Why is a/g
used instead of just a
when doing individual hypothesis tests in the Bonferroni multiple
comparison procedure?
|
[ 11-18-99 ] [ 11-16-99 ] [ 11-11-99 ] [ 11-9-99 ] [ 11-4-99 ] [ 11-2-99 ] [ 10-26-99 ] [ 10-21-99 ] [ 10-19-99 ] [ 10-14-99 ] [ 10-12-99 ] [ 10-7-99 ] [ 10-5-99 ] [ 9-30-99 ] [ 9-23-99 ] [ 9-21-99 ] [ 9-16-99 ] [ 9-14-99 ] [ 9-9-99 ] [ 9-7-99 ] [ 9-2-99 ] [ 8-31-99 ] |
