Statistics II, Midterm, April 22 , 2021 (Test A)

1 Suppose we want to test the hypothesis H 0 ：μ = 20 against H a ：μ ≠ 20. A sample of 100 items will be
taken and the population standard deviation is σ= 10.0. Compute the p-value if the sample mean is
̅
X= 22.0

n > 30 , and σ is known , so we can use Z-test

22−20
with two-sided , p-value = 2*P ( Z > ) = 2*P (Z > 2.0) = 2*(1 – 0.9772) = 0.0456
10/√100

2 Suppose we want to test the hypothesis H 0 ：μ ≥ 10 against H a ：μ < 10. The sample size is 64 and
the sample standard deviation is S = 6.4. Compute the p-value if the sample mean is ̅
X= 8.16

σ is unknown , but n > 30 , so we can use Z-test

8.16−10
with lower-tail , p-value = P(Z < ) = P(Z < -2.3) = P(Z > 2.3) = 1 - 0.9893 = 0.0107
6.4/√64

3 Suppose we want to test the hypothesis H 0 ：μ = 20 H a ：μ ≠ 20. A sample of 121 items will be taken
and the population standard deviation is σ= 7.7. Use α= .05. Compute the probability of making a
Type II error if the population mean is μ= 18.0

n > 30 , and σ is known , so we can use Z-test

7.7
the critical point of rejecting H0 = 20 ± Zα/2*σ/√𝑛 = 20 ± Z0.025* = 18.628 and 21.372
√121

18.628−18 21.372−18
P(tyPe II error) = P(H0 is true | reject H0) = P(Z < ) + P(Z > ) = P(Z < 0.90) + P(Z > 4.82)
7.7/√121 7.7/√121

= 0.1841 + 0 = 0.1841

4 Suppose we want to test the hypothesis H 0 ：p = .20 against H a ：p ≠ .20. A sample of 140 was used.
Compute the p-value when p̅ = .163

0.163−0.2
with two-sided , p-value = 2*P(Z < ) = 2*P(Z > 1.09) = 0.2758
0.2∗0.8
√
140

5 Suppose we want to test the hypothesis H 0 ：p = .23 against H a ：p ≠ .23. A sample of 240 is used.
Compute the p-value of the test when p̅=0.18.

0.18−0.23
with two-sided , p-value = 2*P(Z < ) = 2*P(Z > 1.84) = 0.0658
0.23∗0.77
√
240
6 Young adult magazine states the following hypotheses about the mean age of its subscribers. The
management wants to test the hypothesis H 0 ：μ = 28 against H a ：μ ≠ 28. If the manager conducting
the test will permit a .05 probability of making a Type II error when the true mean age is 30, what
sample size should be selected? Assume σ= 7.2 and a .05 level of significance.

Permitting a 0.05 probability of making a Type II error , so β= 0.05 , Zβ = 1.645

With two-sided , α/2=0.025 , Zα/2=1.96 , σ= 7.2

(|Zα|+|Zβ|)σ 2 (|1.96|+|1.645|)7.2 2
Use the formula : n = ( ) = ( ) = 168.42 , use 169
|μα−μ0| |28−30|

7 A random sample of n=200 is used in testing the hypothesis H 0 ：p = 0.3 vs. H a ：p  0.3. The sample

proportion is p̅ =0.35. What is the p-value of the test？

n > 30 , so we can use Z-test

0.35−0.3
With two-sided , p-value = 2*P(Z > ) = 2*P(Z > 1.54) = 2*(1 - 0.9382) = 0.1236
0.3∗0.7
√
200

8 A random sample of n=200 is used in testing the hypothesis H 0 ：p  0.25 vs. H a ：p > 0.25. The p-
value is 0.0401. What is the sample frequency X？

𝑋
n > 30 , so we can use Z-test p̅= 200

p̅ − 0.25 p̅ − 0.25
p-value = P(Z >
0.25∗0.75
) = 0.0401 → 0.25∗0.75
= 1.75 → X = 60.71 , use 61
√ √
200 200

9 A random sample is to be used in testing the hypothesis H 0 ：  40 vs. H a ： > 40. The significance

level is set at α =0.05 and the population standard deviation of σ=22.0 is known. It is also desired to
be at least 95% sure to reach the correct conclusion when  = 44. What is the sample size n.

σ=22.0 , α= 0.05 , Zα = 1.645 , at least 95% sure → β= 0.05 , Zβ = 1.645

(|Zα|+|Zβ|)σ (|1.645|+|1.645|)22
Use the formula : n = ( |μα−μ0| )2 = ( |40−44| )2 = 327.4 , use 328

10 Out of 180 sampled for the purpose of estimating the population proportion, 83 Successes were
obtained. What is this 95% confidence interval of the population proportion p？

83 p
̅ (1−p
̅) 0.46∗0.54
p̅ = = 0.46 , 95% CI for p：0.46 ± Zα/2*√ = 0.46 ± 1.96*√ = [0.3872 , 0.5328]
180 𝑛 180
Part II (80%) Show all your work!
1(8%) Remote control lawn mowers have been on the sketch board for quite a while but failed to become a
competitive item on the market for various reasons. Marketing experts claimed that as long as the
mean price for the traditional push-and-cut lawn mowers is lower than $350, there is no room for
remote control ones. One investment agency wants to know whether it is time to put money into
remote control lawn mower manufacturing in the year 2020. To test the hypothesis H 0 ：  350 vs.
H a ： > 350, a sample is taken and it is given that the population standard deviation σ=$150. Use a
0.05 significance level. There are two power curves in the following plot. Answer the following
questions according to the plot.
a. One test is obviously better than the other. What is the power of test when the actual mean is $375
if we are using the better test?

We have power curves , so we can just refer the corresponding figure → 0.75

b. What is the Type II risk when the actual mean is $385 if we are using the second best test?

We are asked to calculate the probability of type II error , β

POWER = 1 – β → β = 1 – POWER = 1 – 0.75 = 0.25

2(6%) Young adult magazine states the following hypotheses about the proportion of it’s subscribers that
are over the age of 65. The hypotheses of concern are
H 0 ：p ≥ .14
H a ：p < .14.
If the manager conducting the test will permit a 0.10 probability of making a Type II error when the
true proportion p is 0.09, what sample size should be used? Assume a 0.05 level of significance.

|Zα|√p0(1−p0)+|Zβ|√pa(1−pa) 2 1.645√0.14∗0.86 + 1.28√0.09∗0.91

n=( ) =( )2 = 351.2673
|p0 − pa| 0.14−0.09

→ use 352
3(12%) The global Business Travel Association reported the domestic airfare for business travel for the
current year and the previous year. Below is a sample of 9 flights with their domestic airfares shown
for both years.
Current Year 343 525 414 212 280 409 622 702 589
Previous Year 324 464 459 211 283 429 576 673 596
Di 19 61 -45 1 -3 -20 46 29 -7
The test of concern is
H0 : μD ≤ 0
Ha : μD > 0,
Where μD =μ1 -μ2 . Fill in the following blanks.
The 95% C.I. for μD is [ -16.4813 , 34.4813 ]

The test statistic is 0.8144

The p-value of the test is > 0.1

̅ = 9 , SD = 33.15
D n < 30 , use T-test

(1)
33.15
95% CI for μD ：D
̅ ± t8,0.025* = 9 ± 25.4813 = [ -16.4813 , 34.4813 ]
√9
(2) (3)
9−0
t= = 0.8144 p-value = P(t8 > 0.8144) > 0.1
33.15/√9

We are hard to exactly get the p-value of T-test , so we use a interval to show it.

4(6%) Consider the following hypothesis test.

H0 : μ1 -μ2 =0
Ha : μ1 -μ2 ≠0.
The following results are from independent samples taken from two populations.
Sample1 Sample 2
n= 20 25
X̅ = 900 1000
S= 10 15
What is the degrees of freedom for the t distribution?

s 2 s 2 102 152 2
( 1 + 2 )2 ( + )
n1 n2 20 25
df =
s 2 s 2
= 102 152
= 41.7 , use 41
( n1 )2 ( n2 )2 ( 20 )2 ( 25 )2
1 + 2 +
n1 −1 n2 −1 20−1 25−1
5(6%) Consider the following hypothesis test.
H0 : μ1 -μ2 ≥0
Ha : μ1 -μ2 <0.
The following results are from independent samples taken from two populations.
Sample1 Sample 2
n= 12 16
X̅ = 286 325
S= 40 44
What is the degrees of freedom for the t distribution?

s 2 s 2 402 442 2
( 1 + 2 )2 ( + )
n1 n2 12 16
df =
s 2 s 2
= 402 2 442
= 24.9 , use 24
( n1 )2 ( n2 )2 ( 12 ) ( 16 )2
1 + 2 +
n1 −1 n2 −1 12−1 16−1

6(10%) Periodically, Merrill Lynch customers are asked to evaluate Merrill Lynch financial consultants and
services. Higher ratings on the client satisfaction survey indicate better service, with 7 the maximum
service rating. Independent samples of service ratings for two financial consultants are summarized
here. Consultant A has 10 years of experience, whereas consultant B has 1 year of experience. Use a
= .05 and test to see whether the consultant with more experience has the higher population mean
service rating.
Consultant A Consultant B
n= 16 11
X̅ = 6.74 6.20
S= .66 .70
The null and alternative hypotheses are
H0 : μ1 -μ2 ≤0
When σ1 = σ2 ,
Ha : μ1 -μ2 >0.
we use the pooled estimator
It is assumed that the population variances are equal.
(𝑛1 −1)S1 2 +(𝑛2 −1)S2 2
a. Compute the value of the test statistic. Sp2=
𝑛1 +𝑛2 −2
b. What is the p-value of the test?

(16−1)0.662 +(11−1)0.72
Sp2 = √ = 0.45736
16+11−2

a. b.
(6.74 – 6.2) − 0
t= 1 1
= 2.04 p-value = p(tn1+n2-2 > 2.04)
√0.45736( + )
16 17

→ 0.025 < p-value < 0.05

7(12%) Consider the following hypothesis test.
H0 : μ1 -μ2 =0
Ha : μ1 -μ2 ≠0.
The following results are from independent samples taken from two populations.
Sample1 Sample 2 s 2 s 2
( 1 + 2 )2
n1 n2
n= 20 25 df = s1 2 2 s 2
(n ) ( 2 )2
1 n2
X̅ = 900 1000 n1 −1
+
n2 −1

σ = 10 15
102 152 2
( + )
Fill in the following blanks. = 20 25
102 2 152 2
( ) ( )
20 + 25
The 95% C.I. for μ1 -μ2 is [-107.33 , -92.67] 20−1 25−1

The test statistic is -26.73

= 41.784 , use 41
The p-value of the test is 0
(1)
95% CI for μ1 -μ2 ：
𝑠 2 𝑠2 2
𝑥2 ± t41,0.025*√ 𝑛1 +
̅̅̅1 − ̅̅̅)
(𝑥 (we have no t41,0.025 , so we use the approximation , t31,0.025 )
1 𝑛2

102 152
= (900 - 1000) ± 2.04*√ 20 + = -100 ± 7.33 = [-107.33 , -92.67]
25

(2)
̅̅̅−x
(x 1 ̅̅̅)
2 − 0 (900 − 1000) − 0
t=
𝑠 2 𝑠 2
= = -26.73
√ 1 +2 102 152
𝑛1 𝑛2 √ +
20 25

(3)
p-value = P(t41 < -26.73) ≅ 0

8(10%) Consider the following hypothesis test.

H0 : μ1 -μ2 = 10
Ha : μ1 -μ2 ≠ 10.
It is calculated that the Satterthwaite degrees of freedom is 22, 𝑋̅1 − 𝑋̅2 = −30, and the standard
error of 𝑋̅1 − 𝑋̅2 is 15.95.
a. Compute the value of the test statistic.
b. What is the p-value of the test?
a.
̅̅̅−x
(x 1 ̅̅̅)
2 − 10 −30 − 10
t= = 15.95
= -2.508
standard error

b.
p-value = 2*P(t22 < -2.508) = 2*0.01 = 0.02