Probability and Statistics With R For Engineers and Scientists 1st Edition Michael Akritas Solutions Manual Instant Download All Chapter
Probability and Statistics With R For Engineers and Scientists 1st Edition Michael Akritas Solutions Manual Instant Download All Chapter
Probability and Statistics With R For Engineers and Scientists 1st Edition Michael Akritas Solutions Manual Instant Download All Chapter
https://testbankdeal.com/product/probability-and-statistics-for-
engineers-and-scientists-for-engineers-9th-edition-johnson-
solutions-manual/
https://testbankdeal.com/product/probability-and-statistics-for-
engineers-and-scientists-9th-edition-walpole-solutions-manual/
https://testbankdeal.com/product/essentials-of-probability-
statistics-for-engineers-scientists-1st-edition-walpole-
solutions-manual/
https://testbankdeal.com/product/introduction-to-probability-and-
statistics-for-engineers-and-scientists-5th-edition-ross-
solutions-manual/
Probability With Applications and R 1st Edition Dobrow
Solutions Manual
https://testbankdeal.com/product/probability-with-applications-
and-r-1st-edition-dobrow-solutions-manual/
https://testbankdeal.com/product/probability-and-statistics-for-
engineers-5th-edition-scheaffer-solutions-manual/
https://testbankdeal.com/product/applied-statistics-and-
probability-for-engineers-7th-edition-montgomery-solutions-
manual/
https://testbankdeal.com/product/applied-statistics-and-
probability-for-engineers-6th-edition-montgomery-solutions-
manual/
https://testbankdeal.com/product/statistics-for-engineers-and-
scientists-4th-edition-william-navidi-solutions-manual/
121
Chapter 7
Confidence and Prediction Intervals
40
30
20
Theoretical Quantiles
The Q-Q plot shows that the normal assumption is approximately valid.
2. (a) Using the R commands x=c(649, 832, 418, 530, 384, 899, 755);
confint(lm(x∼1),level=0.9), we find the 90% CI for the true mean histamine
content for all worker bees of this age as (489.9886, 786.2971). In order to make
the CI valid, we need to assume that the data are distributed approximately
normal.
(b) False
√
3. (a) The 80% CI √ for the mean breaking strength μ is X̄ ± tn−1,α/2 S/ n, or 210 ±
t50−1,0.1/2 18/ 50, which is calculated as (206.69, 213.31). Since the sample
size n = 50 is large enough, the normality assumption is not necessary.
(b) Yes
(c) No
4. (a) In Exercise 1, after sorting, the data is 19.62, 36.75, 36.86, 41.72, 46.84, 47.53,
48.42, 48.82, 50.59, 57.16, 62.69. Thus, the CI (36.86, 50.59) is X(3) , X(9) . By
the command 2*(1-pbinom(11-3, 11, 0.5)), we find α = 0.06542969, thus the
confidence level is (1 − α)100% = 93.46%. The CI (36.75, 57.16) is X(2) , X(10)
and the confidence level is 98.83%.
(b) In Exercise 2, after sorting, the data is 384, 418, 530, 649, 755, 832, 899. Thus,
the CI (418, 832) is X(2) , X(6) . By the command 2*(1-pbinom(7-2, 7, 0.5)),
we find α = 0.125, thus the confidence level is (1 − α)100% = 87.5%.
350
300
250
−1 0 1
Theoretical Quantiles
The Q-Q plot shows that the normal assumption is not quite valid.
(b) The 90% CI for the mean ozone level is given by the command
confint(lm(x∼1),level=0.9) and the result is (256.123, 316.5913).
(c) The 90% CI for the median ozone level is given by the command library(BSDA);
SIGN.test(x, alternative=“two.side”, conf.level=0.9) and the result is (248.0466,
291.1626).
(d) The length for the 90% CI for the mean is 60.468, while the length for the 90%
CI for the median is 48.116. Clearly, the 90% CI for the median is shorter.
Since the normal assumption of the data seems not appropriate, we would
prefer the 90% CI for the median.
6. (a) The 95% CI for the mean solar intensity is given by the command
confint(lm(x∼1),level=0.95) and the result is (706.4223, 721.9277). Since the
sample size n = 40 is large enough, the normality assumption is not necessary.
(b) The 90% CI for the median solar intensity is given by the command li-
brary(BSDA); SIGN.test(x, alternative=“two.side”, conf.level=0.95) and the
result is (704.8189, 728.0892). No assumption is needed.
(c) The interpretation of the confidence level is wrong for both the CI in (a) and
(b).
7. (a) For Poisson(λ) distribution, μ = λ. Thus, the 95% CI for λ is the same as the
95% CI for μ. We use the following command:
x=c(rep(0,4), rep(1,12),rep(2,11),rep(3,14),rep(4,9));
confint(lm(x∼1),level=0.95)
to find the CI as (1.888116, 2.591884).
(b) For Poisson(λ) distribution, σ 2 = λ. Thus, the 95% CI for σ is
√ √
( 1.888116, 2.591884), or (1.374087, 1.609933).
8. (a) The 95% CI for the mean eruption duration is given by the command
confint(lm(ed∼1),level=0.95) and the result is (3.351534, 3.624032).
(b) The 90% CI for the median eruption duration is given by the command li-
brary(BSDA); SIGN.test(x, alternative=“two.side”, conf.level=0.95) and the
result is (3.833, 4.1115).
(c) The sample proportion, p̂, can be found with the R command
phat=sum(ed>4.42)/length(ed), which gives 0.2683824. Then using the fol-
lowing commands
alpha=0.05;
phat-qnorm(1-alpha/2)*sqrt(phat*(1-phat)/length(ed));
phat+qnorm(1-alpha/2)*sqrt(phat*(1-phat)/length(ed)),
gives the 95% CI for the probability that an eruption duration will last more
than 4.42 min as (0.2157221, 0.3210426).
9. (a) To find the 95% confidence interval for the proportion, p, of customers who
qualify, we use the following commands:
thus the 95% CI for the true slope of the regression line is
1.55
1.50
1.45
The scatterplot of the data shows that the linearity and homoscedasticity
assumptions of the simple linear regression model seem to be valid.
(b) Using the R commands out=lm(y∼x); out$coef, we have the LSE estimates
α̂1 = 3.130417 and β̂1 = −1.076647. Using the command
sqrt(sum(out$resid**2)/out$df.resid), we find the LSE for σ as 0.0298461.
(c) Using the R command confint(out, level =0.9), we find the 90% CI for β1
as (-1.279482, -0.873812). Since the mean difference of the strengths at wa-
ter/cement ratios 1.55 and 1.35 is μY |X(1.55) − μY |X(1.35) = 0.2β1 , we calculate
the 90% CI for the mean difference of the strengths at water/cement ratios
1.55 and 1.35 as (-0.2558964, -0.1747624).
0.00
−0.02
−0.04
Theoretical Quantiles
The plot does not suggest serious departure from the normality assumption
and thus, the above CIs can be reasonably trusted.
20
15
y
10
5
60 70 80 90
The scatterplot of the data shows that the linearity and homoscedasticity
assumptions of the simple linear regression model seem to be valid.
(b) Using the R commands out=lm(y∼x); confint(out, level =0.95), we find the
95% CI for β1 as (-0.2236649, -0.117264). Using the R commands
t=data.frame(x=80); predict(out, t, interval=“confidence”, level=0.95), we
find the 95% CI for the expected wind speed on an 80F day as (9.082136,
10.11093).
16. Using the command 2*(1-pbinom(30-10, 30, 0.5)), we find α = 0.04277395, thus
the confidence level is (1 − α)100% = 95.72%.
17. Using the command n=16; a=4; 1-2*(1-pbinom(n-a,n,0.5)), we have the result
97.87%; changing a to 5, we have the result 92.32%. These are consistent with
Example 7.3-7.
We find the 95% CI for σ as (0.468561, 1.009343). In order to make the CI valid,
we need to assume that the population distribution is normal.
19. From (7.3.19), we find the (1 − α)100% CI for σ is
n−1 n−1
2
S<σ< 2
S.
χn−1,α/2 χn−1,1−α/2
We find the 95% CI for σ as (0.09463803, 0.1532936). The traditional value of 0.1
lies within the CI.
20. Using the R commands
We find the 95% CI for the population variance of reaction times of Robot 2 as
(0.4916012, 1.696162).
46
44
42
−2 −1 0 1 2
Theoretical Quantiles
4. (a) The expected separation distance between the next cyclist, whose distance
from the roadway center line is 15 feet, and a passing car can be estimated as
(b) A distance of 12 feet is not in the range of X values in the data set, so the
desired PI would not be reliable.