Business Statistics Week 7 Tutorial Solutions

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Business Statistics

Business Statistics Week 7 Tutorial Solutions

Question 7.1
The director of manufacturing at a fabric mill needs to determine whether a new machine is
producing a particular type of cloth according to the manufacturer’s specifications, which indicate
that the cloth should have a mean breaking strength of 30kg. The distribution of the breaking
strength is known to be normal.
A sample of 49 pieces of cloth reveals a sample mean breaking strength of 29.3kg and a standard
deviation of 3.5 kg.
a. Is there evidence that the machine is not meeting the manufacturer’s specifications for mean
breaking strength? Use a 5% level of significance. Ensure that you clearly state your
hypotheses, show ALL steps, ALL your working AND interpret your conclusion in context of
this question.

Let X = breaking strength of the fabric (kg)


𝜎 2
Since X~N (μ σ2) 𝑆𝑜 𝑋̅ ~ 𝑁 (µ, 49 )

Step 1: Hypotheses

Ho: μ =30
H1: μ ≠ 30

Step 2: Test statistic


̅ −µ𝑜
𝑋
σ unknown, use 𝑡 = 𝑠⁄
√𝑛

29.3−30
t calc = 3.5⁄ = -1.40
√49

Step 3: Critical Value


tcrit = t0.025, 48 = 2.009

1
Business Statistics

Step 4: Decision Rule and Decision

Reject H0 if tcalc < -tcrit or tcalc > tcrit

Since -2.009<-1.40<2.009, we cannot reject H0

Step 5: Conclusion

We cannot reject H0 at the 5% level of significance. The sample DOES NOT provide

enough evidence against H0. That is, the mean breaking strength is NOT different from 30 Kg.

The new machine is producing cloth according to the manufacturer’s specifications.

b. Assume that the standard deviation of breaking strength of the fabric for all new machines

is known to be 3.5kg. If you are asked to re-do the test in part (a), what is the critical value
for the test?
Sigma is known, so, the critical value is
Zcrit = Z0.05/2 =Z0.025 = 1.96

Video Solution for Q7.1

Question 7.2

The drug approval process overseen by the FDA in the USA has been subject to complaints
from many sides. It has been criticised
• for being too slow, and keeping potentially life-saving drugs off the market;
• and for being too quick, and accepting information from the pharmaceutical
industry without independent checks.

2
Business Statistics

a. In deciding whether a new drug is safe, what do you think should be the null hypothesis?

b. What should be the alternative hypothesis?

c. With this choice, what is the consequence of a Type I Error?

d. What is the consequence of a Type II Error?

a. Ho: The drug is not safe

b. H1: The drug is safe

c. Type I Error: Rejecting the null hypothesis when it is true. This means allowing an

unsafe drug on the market.

d. Type II Error: Retaining the null hypothesis when it is false. This means doing without a

drug that would have been safe

If your answers in (a) and (b) are reversed, the consequences of the errors are also reversed.
That is:

a. Ho: The drug is safe

b. H1: The drug is not safe

c. Type I Error: Rejecting the null hypothesis when it is true. This means doing without a

drug that would have been safe.

d. Type II Error : Retaining the null hypothesis when it is false. This means allowing an
unsafe drug on the market.

So which probability do you want to keep control of? Is there a difference between the way
to test a new drug on the market, and the way to test a drug that has been used for a long
time and about which safety questions have now been raised?
(this is a discussion type question)

Video solution for Q7.2

3
Business Statistics

Question 7.3

With the recent new packaging legislation, there is a lot of interest at the moment in the proportion
of Australians who smoke. Suppose a survey is conducted on the smoking habits of 5,000 randomly
chosen Australians aged 15 years or older. It is found that 784 of the participants in the survey
currently smoke.

a. Test whether there is evidence at the 5% level that the percentage of Australians aged 15 or
over who smoke is greater than 15%.
Let x= number of Australians aged 15 or over who smoke

p  0.1568
n π = (5000)(0.15) = 750

n(1- π)=(5000)(0.85)= 4250

The assumption of normality is reasonable since they are both >5.

Step 1: Hypotheses

H 0 :   0.15
H1 :   0.15

Step 2: Test statistic

p 
Z is distributed approximately as N(0,1)
 (1   )
n Do
Not RR
0.1568  0.15 Reject
zcalc   1.347 H
0.15(1 0.15) Z
5000 0
zcrit= 1.645
Step 3: Critical Value

Zcrit = Z0.05 = 1.645

Step 4: Decision Rule and Decision

Reject H0 if Zcalc > Zcrit


Since 1.347 < 1.645, we cannot reject H0

Step 5: Conclusion

We CANNOT reject H0 at the 5% level of significance.


The sample DOES NOT provide enough evidence against H0.
That is, the proportion of Australians aged 15 years or older who smoke is not above 15%.

4
Business Statistics

b. What is a Type I error in the context of this question?

A Type I error occurs when we mistakenly reject H0

In the context of this question, we mistakenly state that the proportion of Australians aged 15
years or older who smoke is above 15% when it is not.

OR

We mistakenly conclude that the proportion of Australians aged 15 years or older who smoke
is above 15%.
Video solution for Q7.3

Question 7.4
Eighteen per cent of multinational companies provide an allowance for personal long-distance
calls for executives living overseas. Suppose a researcher thinks that multinational companies
are having a more difficult time recruiting executives to live overseas and that an
increasing number of companies are providing an allowance for personal long-distance calls
to these executives to ease the burden of living away from home. To test this claim, a new
study is conducted by randomly sampling 376 multinational companies. Twenty two per
cent of these surveyed companies are providing an allowance for personal long-distance calls
to executives living overseas.

Has the proportion of multinational companies that provide an allowance for personal
long-distance calls for executives living overseas increased? Use the 1% level of significance.

Let x= number of multinational companies that provide an allowance for personal long-distance
calls for executives living overseas

n =376; p = 0.22; α = 0.01

nπ=376 x 0.18=67.68 and n(1-π)=376(1-0.18)=308.32

Note that both are greater than 5, hence the normal approximation is reasonable.

5
Business Statistics

Step 1: Hypotheses

H 0 :   0.18
H1 :   0.18

Step 2:Test statistic

p 
Z is distributed approximately as N(0,1)
 (1   )
n

0.22  0.18
Z calc   2.02
0.18(1 0.18)
376

Step 3: Critical Value

Zcrit = Z0.01 = 2.327


Do
Not RR
Step 4:Decision Reject
H
Reject H0 if Zcalc > Zcrit
Z
Since 2.02 < 2.327, we cannot reject H0 0
zcrit= 2.327

Step 5: Conclusion

We CANNOT reject H0 at the 1% level of significance.


The sample DOES NOT provide enough evidence against H0.
That is, the proportion of multinational companies that provide an allowance for personal long-
distance calls for executives living overseas in NOT higher than 18%.

The sample evidence shows that the proportion of multinational companies that provide an
allowance for personal long-distance calls for executives living overseas has not increased.

Video Solution for Q7.4

6
Business Statistics

Further practice questions

Question 7.5
A Chinese factory produces mother-of-pearl buttons. The diameter of buttons is assumed to be
normally distributed. If the machinery is working correctly, the mean diameter of the buttons will
be 1.3cm. The variance for that machine is known to be 0.0081 cm2. A sample of 16 buttons is
measured with a laser, and found to have a mean diameter of 1.25cm. Test at the 5% level of
significance, the hypothesis that the mean diameter of the population differs from 1.3cm. , show
ALL steps, ALL your working AND interpret your conclusion in context of this question.

Let X = diameter of mother-of-pearl buttons (cm)


0.0081
Since X~N (μ, 0.0081), 𝑆𝑜 𝑋̅ ~ 𝑁 (µ, )
16

Step 1: Hypotheses
H0: μ=1.3
H1: μ≠1.3

Step 2: Test Statistic


X  0
σ known, use Z 

n

1.25  1.3
Z calc 
0.09
16
Z calc   2.22

7
Business Statistics

Step 3: Critical Value


Zcrit = Z0.025 = 1.96

Step 4: Decision Rule and Decision


Reject H0 if Zcalc < -Zcrit or Zcalc > Zcrit
Since -2.22 < -1.96, we can reject H0

Step 5: Conclusion

We CAN reject H0 at the 5% level of significance. The sample DOES provide


enough evidence against H0. That is, the mean diameter of the buttons is different
from 1.3cm.

The machine is not working properly.

You might also like