Professional Documents
Culture Documents
Chapter 6 Confidence Interval Estimates
Chapter 6 Confidence Interval Estimates
1.
A study is run to estimate the mean total cholesterol level in children 2-6 years of age. A
sample of 9 participants is selected and their total cholesterol levels are measured as follows.
185
225
240
196
175
180
194
147
223
Generate a 95% confidence interval for the true mean total cholesterol levels in adults
with a history of hypertension.
X
185
225
240
196
175
180
194
147
223
1,765
X
X2
34,225
50,625
57,600
38,416
30,625
32,400
37,636
21,609
49,729
352,865
X 1,765
196.1
n
9
X 2 ( 2 )/n
352,865 (1,765) 2 /9
n 1
8
352,865 346,136.1
6728.89
841.1 29.0
8
8
s
Xt
n , df = n 1 = 9 1 = 8
s
2.
145 130
142 135
160
158
4 Weeks
122
142
135
158
155
140
130
120
124
130
Xd
152
155
143
140
126
130
121
120
115
124
Difference = 4 Weeks-Baseline
2
-3
5
-2
3
-3
4
-1
9
-5
9
135
130
Difference2
4
9
25
4
9
9
16
1
81
25
183
Differenc e 9
0 .9
n
10
19.43 4.4
n 1
9
s
Xd t d
n , df = n 1 = 10 1 = 9
For 95% confidence, t = 2.262
s
Xd t d
n
4.4
0.9 (2.262)
10
0.9+3.1
(-2.2, 4.0)
3.
The main trial is conducted and involves a total of 200 patients. Patients are enrolled and
randomized to receive either the experimental medication or the placebo. The data shown below
are data collected at the end of the study after 6 weeks on the assigned treatment.
Experimental
(n=100)
120.2 (15.4)
14%
6%
Placebo (n=100)
131.4 (18.9)
22%
8%
a. Generate a 95% confidence interval for the difference in mean systolic blood pressures
between groups.
1
1
(X1 - X 2 ) ZSp
n1 n 2
2
2
2
2
Check s1 /s2 = 15.4 /18.9 =0.66
Sp
Sp
(n 1 1)s 12 (n 2 1)s 22
n1 n 2 2
1
1
100 100 .
-11.2 + 4.77
(-16.0, -6.4)
b. Generate a 95% confidence interval for the difference in proportions of patients with
hypertension between groups.
(p1 - p 2 ) Z
p1 (1 - p1 ) p 2 (1 p 2 )
n1
n2
Sample size is adequate with at least 5 successes and failures in each group.
100
100
Ln( RR) Z
(n 1 - x 1 )/x 1 (n 2 - x 2 )/x 2
n1
n2
Sample size is adequate with at least 5 successes and failures in each group.
94/6 92/8
Ln(0.75) 1.96
100 100
-0.29+ 1.96 (0.521)
-0.29+1.02
(-1.31, 0.73)
The 95% confidence interval for Ln(RR) is (-1.31, 0.73). To generate the confidence
interval for the relative risk, we take the antilog (exp) of the lower and upper limits:
(exp(-1.31), exp(0.73))
(0.27, 2.08)
4.
The following data are collected as part of a study of coffee consumption among
undergraduate students. The following reflect cups per day consumed:
3
8
X
0
X2
3
4
6
8
2
1
0
2
26
9
16
36
64
4
1
0
4
134
X 26
3.3
n
8
X 2 ( 2 )/n
134 (26) 2 /8
134 84.5
7.07 2.7
n 1
7
7
c. Construct a 95% confidence interval for the mean number of cups of coffee consumed
among all undergraduates.
s
Xt
n , df = n 1 = 8 1 = 7
For 95% confidence, t = 2.365
s
Xt
n
2.7
3.3 (2.365)
8
3.3+ 2.26
(1.0, 5.6)
5.
A clinical trial is conducted comparing a new pain medication for arthritis. Participants
are randomly assigned to receive the new medication or a placebo. The outcome is pain relief
within 30 minutes. The data are shown below.
New Medication
Placebo
Pain Relief
44
21
No Pain Relief
76
99
x 1 (n 1 x 1 ) x 2 (n 2 x 2 )
Sample size is adequate with at least 5 successes and failures in each group.
Ln(2.74) 1.96
1
1
1
1
44 76 21 99
6.
The following data are collected in a clinical trial evaluating a new compound designed
to improve wound healing in trauma patients. The new compound is compared against a
placebo. After treatment for 5 days with the new compound or placebo the extent of wound
healing is measured and the data are shown below.
Treatment
New Compound (n=125)
Placebo (n=125)
Suppose that clinicians feel that if the percent reduction in the size of the wound is
greater than 50% then the treatment is a success.
a. Generate a 95% confidence interval for the percent success in patients receiving the new
compound.
p = 73/125 = 0.584
p(1 - p)
n
Sample size is adequate with at least 5 successes and failures in the sample.
p Z
0.584(1 - 0.584)
125
0.584 + 0.086
0.584 1.960
(0.498, 0.670)
b. Generate a 95% confidence interval for the difference in the percent success between the
new compound and placebo
p1 (1 - p1 ) p 2 (1 p 2 )
n1
n2
Sample size is adequate with at least 5 successes and failures in each group.
125
125
RR = p 1/ p 2 = 0.584/0.352 = 1.66
Ln( RR) Z
(n 1 - x 1 )/x 1 (n 2 - x 2 )/x 2
n1
n2
Sample size is adequate with at least 5 successes and failures in each group.
52/73 81/44
Ln(1.66) 1.96
125
125
0.507 + 1.96 (0.143)
0.507 + 0.28
(0.227, 0.787)
The 95% confidence interval for Ln(RR) is (0.227, 0.787). To generate the confidence
interval for the relative risk, we take the antilog (exp) of the lower and upper limits:
(exp(0.227), exp(0.787))
(1.25, 2.20)
d. Generate a 95% confidence interval for the odds ratio of treatment success between
treatments.
1
1
1
1
x 1 (n 1 x 1 ) x 2 (n 2 x 2 )
Sample size is adequate with at least 5 successes and failures in each group.
Ln(2.59) 1.96
1
1
1
1
73 52 44 81
The 95% confidence interval for Ln(OR) is (0.44, 1.46). To generate the confidence
interval for the odds ratio, we take the antilog (exp) of the lower and upper limits:
(exp(0.44), exp(1.46))
(1.55, 4.31)
7.
p1 (1 - p1 ) p 2 (1 p 2 )
n1
n2
Sample size is adequate with at least 5 successes and failures in each group
.
0.38(1 0.38) 0.21(1 0.21)
(0.38 0.21) 1.96
100
100
0.17 + 1.96 (0.063)
0.17 + 0.12
(0.05, 0.29)
b. Estimate the relative risk for reduction in symptoms between groups
RR = p 1/ p 2 = 0.38/0.21 = 1.81
Ln( RR) Z
(n 1 - x 1 )/x 1 (n 2 - x 2 )/x 2
n1
n2
Sample size is adequate with at least 5 successes and failures in each group.
Ln(1.81) 1.96
62/38 79/21
100
100
8.
The following table displays descriptive statistics on the participants involved in the
study described in Problem 7.
Characteristic
Mean (SD) Age, years
% Males
Mean (SD) Educational Level, years
Mean (SD) Annual Income, $000s
Mean Body Mass Index
Experimental Medication
(n=100)
47.2 (4.3)
46%
13.1 (2.9)
$36.560 ($1,054)
24.7 (2.7)
Placebo
(n=100)
46.1 (5.1)
58%
14.2 (3.1)
$37.470 ($998)
25.1 (2.4)
a. Generate a 95% confidence interval for the mean age among participants assigned to the
placebo.
s
XZ
n
5.1
46.1 (1.960)
100
46.1 + 0.999
(45.1, 47.1)
b. Generate a 95% confidence interval for the difference in mean ages in participants
assigned to the experimental versus the placebo groups.
1
1
(X1 - X 2 ) ZSp
n1 n 2
2
2
2
2
Check s1 /s2 = 4.3 /5.1 =0.71
Sp
Sp
(n 1 1)s 12 (n 2 1)s 22
n1 n 2 2
22.25 4.72
100 100 2
(47.2 - 46.1) 1.960(4.72)
1
1
100 100 .
1.1 + 1.31
(-0.21, 2.41)
c. Generate a 95% confidence interval for the difference in mean body mass index in
participants assigned to the experimental versus the placebo groups.
(X1 - X 2 ) ZSp
1
1
n1 n 2
Sp
(n 1 1)s 12 (n 2 1)s 22
n1 n 2 2
6.52 2.55
1
1
100 100 .
(-1.11, 0.31)
9.
sd
n
- 9.4 (1.960)
2.8
50
-9.4 + 0.78
(-10.2, -8.6)
10.
Suppose in the study described in Problem 9, each participantis also asked if the assigned
medication causes any stomach upset. Among the 50 participants, 12 report stomach upset with
the experimental medication. Construct a 90% confidence interval for the proportion of
participants who might experience stomach upset with the experimental medication.
p = 12/50 = 0.24
p(1 - p)
p Z
n
Sample size is adequate with at least 5 successes and failures in the sample.
0.24(1 - 0.24)
50
0.24 + 0.099
0.24 1.645
(0.14, 0.34)
11.
A 95% confidence interval for the mean diastolic blood pressure in women is 62 to 110.
What is the margin of error? What is the standard error?
CI has form: 86 + 24
Margin of error is 24 (1.96*Standard Error = 24), thus Standard Error = 12.24.
12.
The following data are collected from 10 randomly selected patients undergoing physical
therapy following knee surgery. The data represent the percent gain in range of motion
after 3 weeks of therapy. Generate a 95% confidence interval for the mean percent gain
in range of motion.
24%
32%
50%
62%
21%
X
45%
378
10
80%
24%
30%
10%
= 37.8
X 2 (X) 2 /n
18,386 (378) 2 /10
18,386 14,288.4
455.29 21.3
n 1
10 1
9
Xt
s
n , df = n 1 = 10 1 = 9
The following data are collected in a randomized trial to test the efficacy of a new drug
for migraine headaches. The following are characteristics of study participants overall
and then organized by the treatment to which they are assigned.
New Drug
(n=100)
32.8 (4.7)
54%
Placebo
(n=100)
31.9 (5.1)
48%
All
(n=200)
32.0 (4.9)
51%
76%
80%
78%
22%
38%
40%
20%
42%
38%
21%
39%
39%
5 (3-12)
0-35
6 (2-18)
0-48
6 (3-17)
0-48
a. Generate a 95% confidence interval for the difference in mean ages between groups.
(X1 - X 2 ) ZSp
1
1
n1 n 2
Sp
(n 1 1)s 12 (n 2 1)s 22
n1 n 2 2
24.05 4.90
1
1
100 100 .
0.9 + 1.36
(-0.46, 2.26)
b. Generate a 95% confidence interval for the difference in proportions of males
between groups.
(p1 - p 2 ) Z
p1 (1 - p1 ) p 2 (1 p 2 )
n1
n2
Sample size is adequate with at least 5 successes and failures in each group.
100
100
0.06 + 0.14
(-0.08, 0.20)
c. Generate a 95% confidence interval for the difference in proportions of patients with
severe migraine headaches between groups.
(p1 - p 2 ) Z
p1 (1 - p1 ) p 2 (1 p 2 )
n1
n2
Sample size is adequate with at least 5 successes and failures in each group.
100
100
A clinical trial is run to assess the efficacy of a new pacemaker device in patients with
atrial fibrillation (AF). Two hundred participants are randomized to receive the new
pacemaker or a currently available pacemaker. There are two primary outcomes of
interest the number of days in a three month period with an atrial fibrillation event and
hospitalization for atrial fibrillation over the three month follow-up period. Data on
baseline characteristics and the outcomes are shown below.
Baseline Characteristics
Mean (SD) Age, years
% Male
Outcomes
Mean (SD) Number of days with AF
event
% Hospitalized for AF
a.
New Pacemaker
(n=100)
67.3 (5.9)
48%
Available Pacemaker
(n=100)
66.9 (5.6)
52%
8.4 (3.2)
14.9 (3.9)
4%
9%
Compute a 95% confidence interval for the difference in mean number of days with
an AF event between participants receiving the new pacemaker as compared to the
availablepacemaker.
1
1
n1 n 2
(X1 - X 2 ) ZSp
Check s12/s22 = 3.22/3.92 = 0.67
Sp
Sp
(n 1 1)s 12 (n 2 1)s 22
n1 n 2 2
12.73 3.56
100 100 2
1
1
100 100 .
Compute a 95% confidence interval for the mean number of days with an AF event
among participants receiving the new pacemaker.
XZ
s
n
8.4 (1.960)
3.2
100
8.4 + 0.63
(7.77, 9.03)
15.
A pilot study is run to investigate the feasibility of recruiting pregnant women into a study of risk
factors for preterm delivery. Women are invited to participate at their first clinical visit for
prenatal care. The following represent the gestational ages in weeks of women who consented to
participate in the study. Compute a 95% confidence interval for the mean gestational age of
women enrolling in the study.
11
14
21
22
118
8
= 14.8
10
13
18
X 2 (X) 2 /n
1916 (118) 2 /8
1916 - 1740.5
25.07 5.0
n 1
8 1
7
Xt
s
n , df = n 1 = 8 1 = 7
A clinical trial is run comparing a new drug for high cholesterol to a placebo. A total of
40 participants are randomized (with equal assignment to treatments) to receive either the
new drug or placebo. Their total serum cholesterol levels are measured after eight weeks
on the assigned treatment. Participants receiving the new drug reported a mean total
serum cholesterol level of 209.5 (std dev = 21.6) and participants receiving the placebo
reported a mean total serum cholesterol level of 228.1 (std dev = 19.7).
a. Construct a 95% confidence interval for the difference in mean total serum
cholesterol levels between participants receiving the new drug versus placebo.
1
1
n1 n 2
df = n1 + n2 -2 = 20 + 20 -2 = 38
(X1 - X 2 ) tSp
Sp
(n 1 1)s 12 (n 2 1)s 22
n1 n 2 2
427.33 20.7
1
1
20 20 .
-18.6 + 13.25
(-31.85, -5.35)
b. Is the new drug effective? Justify briefly.
Yes. The 95% confidence interval for the difference in means does not include the
null value (0) and thus there is a statistically significant difference in mean total
serum cholesterol levels between participants receiving the new drug versus placebo.