STAT 426 HW4.
1. (2.21)
Let
u
= (1
,
2
,
3) for women’s attitudes (not likely, somewhat likely, very
likely), and
v
= (0
,
2
,
1) for mammography experience (never, over one year
ago, within the past year).
From hw3, we know that 22% cells have expected counts less than 5; there
fore, we prefer Fisher’s Exact test to the ChiSquare test.
From SAS output, pvalue= 2
.
52
×
10

5
. Therefore, mammography experi
ence and women’s attitudes are associated.
The standardized residuals are
Not likely Somewhat likely Very likely
Never
1.35
3.96
4.67
>
1 year
0.48
0.84
0.58
≤
1 year
1.96
3.76
4.00
Women who think it is very likely a mammogram will find breast cancer
are more likely to have had an examination, whereas those who think the
examination is only somewhat effective or not very effective are more likely
not to have had an examination.
2. (2.25)
Let
θ
be the ratio of the odds of cancer controlled by surgery to the odds of
cancer controlled by radiation therapy.
pvalue=
23
∑
m
=21
parenleftBig
23
m
parenrightBigparenleftBig
18
36

m
parenrightBig
parenleftBig
41
36
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
This is the end of the preview.
Sign up
to
access the rest of the document.
 Spring '10
 Xe
 ChiSquare Test, Statistical tests, Chisquare distribution, Fisher's exact test, exact test

Click to edit the document details