Test 1, Statistics, 4/1/2021 B

Part I (12%): Calculate p-values. Just refer the probability table

Test using N(0, 1) Test using t-dist
(σ is known or n≥30) (σ is unknown and n < 30)
Type of test Test statistic p-value Type of test Test Degrees of Range of p-
Z* statistic t* freedom value
Lower-tail -1.89 0.0294 Upper-tail 1.93 17 [0.025,0.05]
Two-sided -2.11 0.0348 Two-sided -2.02 11 [0.05,0.1]
We should double the p-value when Type of test is Two-sided
Part II(68%) Four points each.

1 Let X~N(0, 1), and P(X< a)=0.3594, P(X > b)=0.1271, then a=_-0.36_, and b=_1.14_.
0.3594 + 0.5 = 0.8594 , Z0.8594 = 0.36 → Z0.3594 = -0.36 Z0.1271 = 1.14

2 Let X~N(0, 1), then P(X <−0.25)=_0.4013_, P(-0.42  Z  1.06)=_0.5182_.

The question asks us P(X <−0.25) , but we only have a table which X > 0 , therefore we must refer the
symmetric figure and take “1” minus it ( Z(0.25) = 0.5987 , 1 - 0.5987 = 0.4013 )

P(-0.42  Z  1.06) = P( Z  1.06) – P(Z  -0.42) = 0.8554 – 0.3372 = 0.5182

3 Let X~ N(20, 81), then P(X >15.5)=_0.6915_, P(11 ≤ X ≤ 24.5)=_0.5328_.

Because X is not a situation in =0 ,σ=1 , we can use probability table only if standardize X in advance

15.5−20
P(X >15.5) = P(Z > ) = P(Z > -0.5) = 1 – P(Z < 0.5) = 0.6915
9

24.5−20 11−20
P(11 ≤ X ≤ 24.5) = P(Z < ) - P(Z < ) = 0.5328
9 9

4 Let N(0, 1), and P(-a < X< a)=0.2206, P(-b < X < b)=0.9070, then a=_0.28_, and b=_1.68_.

Since  = 0 , we can confirm –X and X are mutually symmetric , P(-a < X< a) = 2 x P (0 <X< a)
= 2 x [P(X < a) - 0.5] = 0.2206 → P(X < a) = 0.6103 → Z0.6103 = 0.28

For the same reason , P( X < b) = 0.9535 , Z0.9535 = 1.68

5 Suppose we want to estimate the population mean . It is given that the population distribution is
approximately normal with a standard deviation =8.5. A random sample gave n=25, and X ＝42.0.
What is the 95% confidence interval of this ？
 8.5
95% CI of  : X̅ ± Zα/2 = 42 ± 1.96 * = [38.67,45.33]
√𝑛 √25

6 Out of 180 sampled for the purpose of estimating the population proportion, 72 Successes were
obtained. What is this 95% confidence interval of the population proportion p？

𝑝̅ (1−𝑝̅ ) 0.4𝑥0.6
We have : n=180 , 𝑝̅ = 72/180 = 0.4 → 95% CI = 𝑝̅ ± Zα/2√ = 0.4±1.96√ = [0.3284,0.4716]
𝑛 180
7 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=14.0 is known. It is also desired
to be at least 95% sure to reach the correct conclusion when  = 43.5. Calculate the sample size n.

at least 95% sure →𝛽 = 0.05 , Zβ = 1.645 , Upper-tail → use 𝛼 = 0.05 , Z𝛼 = 1.645

(|Z𝛼|+|Z β|)σ 2 (|1.645|+|1.645|)∗14 2

Use the formula : n = ( ) = ( ) = 173.1856 → use 174
|𝛼−0| |20−23.5|

8 A random sample of n=100 is used in testing the hypothesis H 0 ： = 20 vs. H a ：  20. The sample
mean and standard deviation turned out to be X =17.40 and S=13.0. What is the p-value of the
test？
n > 30 , and σ is known , so we can use z-test
17.4−20
H0： = 20 vs. Ha：  20 → Two-sided → p-value = 2*P(Z < ) = 2*0.2275 = 0.455
13/√100

9 The sample size n is to be determined for the one of the late October poll of the Presidential Election.
The proposed 95% confidence interval should be within 3.3%. Calculate the sample size n.
𝑍𝛼/2 ∙0.5 2 1.96∗0.5 2
Please use the formula : n = ( 𝐸
) →n= ( 0.033
) = 881.91 , use 882

10 A random sample of n=64 is used in testing the hypothesis H 0 ：  40 vs. H a ： > 40. The p-value
of the test turns out to be 0.1003. If the population standard deviation is σ=16.0, what is X ?

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

X̅ −40
H0： ≤ 40 vs. Ha： > 40 → Lower-tail → p-value = P( Z > 16/√64) = 0.1003
X̅ −40
→ Z0.8997 = 1.28 → 16/√64 = 1.28 → X̅ = 42.56

11 A random sample of n=25 drawn from a normal population is used in testing the hypothesis H 0 ： 
20 vs. H a ： < 20. The sample mean turned out to be X =18.0. Suppose it was given that the
population standard deviation is σ=4.5, what is the p-value of the test？

n < 30 , but we have a normal population hypothesis and σ is known , so we can use z-test

18−20
H0：  20 vs. Ha： < 20 → Lower-tail → p-value = P( Z < 4.5/√25) = 0.0132

12 A random sample is to be used in testing the hypothesis H 0 ： = 20 vs. H a ：  20. The significance
level is set at α =0.05 and the population standard deviation of σ=6.0 is known. It is also desired to
be at least 90% sure to reach the correct conclusion when  = 23. Calculate the sample size n.

at least 90% sure → 𝛽 = 0.10 , Zβ = 1.28 , Two-sided → use 𝛼/2 = 0.025 , Z𝛼/2 = 1.96

(|Z𝛼/2|+|Z β|)σ 2 (|1.96|+|1.28|)∗6 2

Use the formula : n = ( ) = ( ) = 41.9904 → use 42
|𝛼−0| |20−23|
13 A random sample is to be used in testing the hypothesis H 0 ：  20 vs. H a ： < 20. The significance
level is set at α =0.05 and the population standard deviation of σ=9.0 is known, a sample of n=81 is
used. What is the probability of making the correct decision when  = 18.5?

n > 30 , and σ is known , so we can use z-test In this situation , we need calculate 1-β
9
H0：  20 vs. Ha： < 20 → Lower-tail → critical value = 20 – 1.645*√81 = 18.355

18.355−18.5
That is , we reject H0 if X̅ < 18.355 , β = P (not rejected H0 | Ha is true) = P (Z > 9/√81 ) = 0.5577

1-β = 1 - 0.5577 = 0.4423

14 A random sample of n=210 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=14.0 is given.
Calculate the Type II risk when the actual mean is  = 42.8
P(Type II risk) = P(𝛽) → First , find the critical value rejected H0
14
critical value : we reject H0 if X̅ out of 40 ± 1.96* = [38.106 , 41.894]
√210

38.106−42.8 41.894−42.8
P(𝛽) = P(not rejected H0 | Ha is true) = P(38.106 < X̅ < 41.894 | = 42.8) = P( 14/ 210 < 𝑍 < 14/√210
)
√

−4.694 -0.906
= P(14/√210 < 𝑍 < ) = P(-4.86 < Z < -0.94) ≈ P(Z < -0.94) = 0.1736
14/√210

15-16 The weight of freshly picked, sun-sweet southern Taiwan Grade A mango is distributed normally
with mean 380g and standard deviation 8.0g.
15 Eight mangos are packed in one medium size Gift Basket. What is the probability that the net weight of
such a basket of mangos being over 3050 g? Please note that the weight of the basket itself is
ignored.
381.25−380
We have : X̅ = 3050/8 = 381.25 , P(Z > ) = P(Z > 0.44) = 0.3300
8/√8
16 There are 12 mangos in one Deluxe Gift Basket. What is the probability that the mean weight of such a
basket of mangos comes between 376g and 385g?
376−380 385−380
P(376 < X < 385) = P( <Z< ) = P(-1.41 < Z < 1.77) = 0.9616 – 0.0793 = 0.8823
8/√8 8/√8

17 A random sample of n=180 is used in testing the hypothesis H 0 ：p = 0.31 vs. H a ：p  0.31. The sample
frequency is X=64. What is the p-value of the test？

p(1−p) 0.31∗0.69
We have : 𝑝̅ = X/n = 64/180 = 0.3556 , standard deviation of p = √ 𝑛 = √ 180 = 0.345

0.3556−0.31
p-value(Two-sided) = 2*P(Z > 0.345
) = 2*0.0934 = 0.1868
18 An examination of the records for a random sample of 11 motor vehicles in a large fleet shows the
following operating costs (in cents per mile):
35.3, 36.4, 31.2, 32.5, 34.5, 32.1, 36.8, 36.7, 33.9, 31.6, 34.8
The statistics were calculated and listed below.
X̅ = 34.16, S = 2.651.
Let  denote the mean operating cost in cents per mile for the vehicles in the fleet. The distribution
of vehicle operating costs is approximately normal. Test the hypothesis H 0 ：  36.0 against H a ： 
36.0 at ＝0.05 level. What is the test statistic.

X̅ − 34.16−36
t* = = = -2.30
𝑠/√𝑛 2.651/√11

19 A random sample of n=300 is used in testing the hypothesis H 0 ：p = 0.25 vs. H a ：p  0.25. The p-value

is 0.0672. What is the sample frequency X？

H0：p = 0.25 vs. Ha：p  0.25 → Two-sided , p-value = 0.0672 → p-value of one side = 0.0336
𝑋
| −0.25|
300
→ P(Z > 0.25∗0.75
) = 0.0336 → X = 88.725 or 61.275 → use 89 and 62
√
300

20 Suppose we want to calculate a 99% confidence interval for the population mean . It is given that the
population distribution is approximately normal with a standard deviation =12.8. A margin of error
no larger than 4.0 is desired. Calculate the sample size n.
𝛼 = 1 - 99% = 0.01 Z 𝛼/2 = 2.575 ,
12.8
A margin of error no larger than 4.0 is desired → 2.575* ≤4 → n ≥ 67.8976 → use 68
√𝑛

21 Miss Beckinsale ordered a huge batch of tiles. It was agreed that the defective rate should be no higher
than 7%. The tiles are here for inspection now. Miss B wanted to be sure that she can reject the tiles
with 95% confidence if the real defective rate is 18%. How many tiles should be inspected?
|Z𝛼|√p0(1−p0)+|Zβ|√pa(1−pa) 2 1.96√0.07∗0.93+1.645√0.18∗0.82 2
n=( ) =( ) = 105.9 → use 106
|p0 − pa| 0.18−0.07
Part III(24%)
1(12%) The Vice President of a community college believes that the college students who rented apartments off
campus spent, on the average, $2,300 per month. A survey with a random sample of 81 students who live
off campus showed a sample mean of $2,406, and a sample standard deviation of $477. It is desired to test
if the average monthly rent is higher than $2300.
a. Test the hypothesis at 0.01 level. That is, calculate the test statistic, specify the action limit, or, the
rejection region, and finally state your conclusion.
b. Compute the p-value of the test.
First , write down the H0 and Ha , we always choose the desired one as Ha
H0 :  ≤ 2300 ,Ha :  > 2300 ,
n = 81 , 𝑥̅ = 2406 ,  = 477 , 𝛼 = 0.01
a:
477
We reject H0 if 𝑥̅ > 2300 + 2.326* = 2423.278 → 𝑥̅ = 2406 < 2423.278 → don’t reject H0
√81

b:
2406−2300
p-value = P(Z > ) = P(Z > 2) = 0.0228
477/√81

2(12%) A federal funding program is available to low-income neighborhoods. To qualify for the funding, a
neighborhood must have a mean household income of less than $15,000 per year. Neighborhoods
with mean annual household income of $15,000 or more do not qualify. Funding decisions are based
on a sample of residents in the neighborhood. A hypothesis test with a 0.02 level of significance is
considered. Use σ=$4,000 as a planning value.
a. If a sample of n=120 residents of a neighborhood is used for the test, what is the probability of funding
a neighborhood with a mean annual household income of $14,000?
b. If the funding guidelines call for a maximum probability of 0.05 of not funding a neighborhood with a
mean annual household income of $14,000. What sample size should be used in the funding decision
study?
First , determine the H0 and Ha , we always choose the desired one as Ha
H0 :  ≥ 15000 , Ha :  < 15000 , n = 120 ,  = 4000 , 𝛼 = 0.02 , Z 𝛼 = 2.05
Second , determine the critical value
4000
We reject H0 if 𝑥̅ < 15000 – 2.05* = 14251.45
√120

a.
Calculate POWER(1-𝛽)
14251.45−14000
1-𝛽 = 1 – P (𝑥̅ > 14251.45| =14000 ) = 1 – P(Z > ) = 1 – 0.3114 = 0.6886
4000/√120

b.
at least 95% sure →𝛽 = 0.05 , Zβ = 1.645 , Lower-tail → use 𝛼 = 0.02 , Z𝛼 = 2.056

(|Z𝛼|+|Z β|)σ 2 (|2.05|+|1.645|)∗4000 2

Use the formula : n = ( ) = ( ) = 218.4484 → use 219
|𝛼−0| |14000−15000|