Exercise 6- Statistical Inferences

Group 5

Problem 1
Weather forecasters would like to report on temperatures in Washington, DC. In particular, they
would like to compare the temperatures in December to temperatures in January to help potential
tourists understand whether there is a difference. They believe it is colder in January but want to
conduct a test to support their assertion. The average December temperature in a random sample
of 16 days is 35 and the average January temperature in a random sample of 20 days is 33. The
standard deviations based on historic information are 4 for December and 5 for January.

January December
x̄ 1 = 35 x̄ 2 = 33
n = 16 n = 20
σ1 = 4 σ2 = 5

m
er as
a. Specify the competing hypotheses to determine if it is colder in January than in
December.

co
eH w
Right-tailed test
H0 ; μ1 − μ2 ≤ d0

o.
HA ; μ1 − μ2 > d0
rs e
b. Find the critical value(s) of the appropriate test at the 5% significance level.
ou urc
c. Calculate the value of the test statistic.
Since the population standard deviations are known, we compute the value of the test
statistic as:
o
aC s
vi y re

(35−33 ) −0
2
Z= 4 2 52 = = 1.33
√ +
16 20
1.5
ed d
ar stu

The p-value of the right-tailed test is computed as p-value = P(Z ≥ 1.33) = 1 – 0.91759 =
0.08241
d. Make a conclusion at the 5% significance level.
We can’t reject the null hypothesis since the p-value of 0.08241 is more than the chosen α
is

= 0.05. Therefore, at the 5% significance level, the weather forecasters can’t conclude
Th

that average temperature in January is colder than that of December.

Problem 2
A large city installed LED streetlights, in part, to help reduce crime in particular areas. The city

This study source was downloaded by 100000775214354 from CourseHero.com on 11-07-2021 22:21:01 GMT -06:00

https://www.coursehero.com/file/114933022/Exercise-6-for-studentsdocx/
 collected the number of property crimes along nine randomly selected streets in the month before
and after the LED light installation, and claims that crime has been reduced. The crime data is
shown in the table.
Before LED After LED
17 14
25 20
10 14
14 12
16 8
34 35
32 27
21 18
14 12

m
er as
a. Specify the competing hypotheses to test the city's claim.
Left-tailed test

co
eH w
H0 ; μD ≥ d0
HA ; μD < d0

o.
b. Calculate the value of the relevant test statistic and find the critical value at the 5%
significance level. rs e
ou urc
Before After
di (di – d)2
LED LED
17 14 3
o

400
25 20 5 324
aC s

10 14 -4
vi y re

729
14 12 2 441
16 8 8 225
34 35 -1 576
ed d

32 27 5 324
ar stu

21 18 3 400
14 12 2 441
Σdi =23 Σ(di – d)2 = 3860
is
Th

Sample mean:
sh

23
d= =2.55
9

This study source was downloaded by 100000775214354 from CourseHero.com on 11-07-2021 22:21:01 GMT -06:00

https://www.coursehero.com/file/114933022/Exercise-6-for-studentsdocx/
 Sample variance:

3860
S2D = = 167. 27 SD = √ 167.27 = 12.93
23−1
We compute the value of the tdf test statistic with df = n − 1 = 9 − 1 = 8 as

2.55−0
t8 = = 0.25
12.93−√ 9

m
er as
c. Does the evidence support the city's claim at the 5% significance level?
Given a right-tailed hypothesis test with df = 8, we can use the t table to approximate the

co
eH w
p-value = P(T8≤ 0.25), as 0.05 < p-value < 0.10. Using statistical software, we find that
the exact p-value = 0.4044. Since the p-value > 0.05, we do not reject H0. At the 5%

o.
significance level, we cannot conclude that a month after the installation of LED the
crime has been reduced rs e
ou urc
Problem 3
The supervisor of an automobile sales and service industry would like the variance of the number
o

of automobiles serviced not to exceed five units-squared. He would recruit one more service
aC s

executive if the variance exceeds this threshold. In a recent random sample of 14 services, he
vi y re

computed the sample standard deviation as 3.3. He believes that the automobile services are
normally distributed.
a. State the null and the alternative hypotheses to test if the threshold has been crossed.
ed d

b. Use the critical value approach to conduct the test at α = 0.01.

Problem 4
is

The sales price (in $1,000) of three-bedroom apartments in two cities—Houston, Texas, and
Th

Orlando, Florida—are given in the following table. Houston is known to have lower house prices
than Orlando; however, it is not clear if it also has lesser variability in its prices.

Houston 189 145 220 205 225 275 248 310 150 318
sh

Orlando 301 279 224 400 249 205 289 350 218 300

a. State the null and the alternative hypotheses to determine if the variance of house price in
Houston is less than that in Orlando.
b. What assumption regarding the population is necessary to implement this step?

This study source was downloaded by 100000775214354 from CourseHero.com on 11-07-2021 22:21:01 GMT -06:00

https://www.coursehero.com/file/114933022/Exercise-6-for-studentsdocx/
 c. Test the hypothesis at α = 0.05. What is your conclusion?

m
er as
co
eH w
o.
rs e
ou urc
o
aC s
vi y re
ed d
ar stu
is
Th
sh

This study source was downloaded by 100000775214354 from CourseHero.com on 11-07-2021 22:21:01 GMT -06:00

https://www.coursehero.com/file/114933022/Exercise-6-for-studentsdocx/
Powered by TCPDF (www.tcpdf.org)