Effect of Growth-promoting Drug on the Weights of Chickens
Block |
Control |
Low Dose |
High Dose |
1
|
3.93
|
3.99
|
3.96
|
2
|
3.78
|
3.96
|
3.94
|
3
|
3.88
|
3.96
|
4.02
|
4
|
3.93
|
4.03
|
4.06
|
5
|
3.84
|
4.10
|
3.94
|
6
|
3.75
|
4.02
|
4.09
|
7
|
3.98
|
4.06
|
4.17
|
8
|
3.84
|
3.92
|
4.12
|
A randomized block experiment was performed to investigate the effect of a potentially growth-promoting drug added to the feed of chickens.
The experimental unit was a group of chickens, reared and fed together in the birdhouse. The experimental units were grouped three to a
block, with physically adjacent units being put into the same block. The response values shown in the table above are the average weights
of the chickens (at maturity) in each experimental unit.
StatXact instructions for Page's test
The four columns of values (the block ID and the three columns of average weight values)
can be pasted into Var1 through Var4 in StatXact's CaseData editor. Then one can use
Nonparametics > K Related Samples > Page...
One can highlight Var2, Var3, and Var4 in the Variables box and click the arrow to put
them into the Populations (Treatments) box. Under compute select Exact, and then click
OK.
The output shows that the observed valued of L is 109, and the exact p-value resulting
from a one-tailed test is about 0.00035.
Notes:
- StatXact's asymptotic p-value does not make use of the continuity correction shown in the text book and in the class notes.
- For a given ordering of treatments, StatXact will consider both one-tailed tests and report the smaller of the two
pertinent p-values. (E.g., if the treatments are entered in the order
A,
B,
and C, StatXact will consider both
μA <=
μB <=
μC
and
μA >=
μB >=
μC,
and report the smaller of the two one-tailed p-values. If you want the p-value for the other ordering, subtract the
given one-sided p-value from 1 and add the given point probability. (I.e., if StatXact gives P(L >= l),
and you want P(L <= l), use 1 -
P(L >= l) +
P(L = l)
= P(L < l)
+ P(L = l)
= P(L <= l).))