Homework 5 solutions
6.1 

Patient number Random digit treatment assignment 
1 6 b 
2 9 b 
3 6 b 
4 4 a 
5 4 a 
6 3 a 
7 7 b 
8 1 a 
9 9 b 
10 8 b 
11 0 a 
12 0 a 
13 0 a 
14 2 a 
15 8 b 
16 9 b 
17 8 b 
18 1 a 
19 9 b 
20 5 b 

6.2 A group = 9 B group = 11, 10 A and 10 B expected 
6.7 .080 

6.8 .015 

6.10 2.583 

6.11 -1.313 

6.12 2.365 

6.13 8.6+ or - 2.36 (6.24, 10.96) 

6.14 7.84 + or - 1.33 (6.51, 9.17) 

6.15 7.84 + or - 1.10 (6.74, 8.94) 

6.16 The 90% CI should be shorter than the 95% CI since we are requiring
less confidence. 

6.44 .013 

6.45 .013 + or - .010 (.0035, .023) 

6.46 incidence is .0040. It falls within the 95% CI so we conclude that
the
cancer rate of Michigan nurses is not different from the connecticut
female
population. 

6.51 .2 

6.52 95% CI=(.122, .278). .10 not in that interval, so pretty confident
that it’s
not a placebo effect. 

6.53 that the normal aprox o to the binomila distribution is
appropriate. It is
because npq=100(.2)(.8)=16 which is greater than 5. 

6.54 1.2 

6.55 (2.9, 7.7) 

6.56 We can conclude that the drug is effective because 0 is not included
within the 95% CI. 

6.57 The 95% CI menas that if a large # of samples of 100 hypertensive
patients were selected and the preceeding type of interval constructed,
then approximately 95% of such intervals would contain the underlying
mean differences in blood pressure upon using the drug. 

6.88 A false positive mamogram is one that is interpreted as positive for
breast cancer when the woman actually does not have breast cancer. 

6.89 95% CI = (.066, .090) ; therefore mamogram is cost effective. 

6.90 p =.334 

6.91 95% CI = (.313, .354)