Statistics – Hypothesis Testing

Lecture Problems
Instructor: Ling-Chieh Kung
Department of Information Management
National Taiwan University

1. You have been selling a product at $50 per unit for one year. It seems to you that the demand is
higher than the supply, so you are considering whether you should increase the unit price. In order
to understand how much a consumer is willing to pay for this product, you randomly sampled 1000
consumers and asked this question. If the average willingness-to-pay is around $50, you will not
change the price. You will increase your price if and only if the average willingness-to-pay is higher
than $60 (not $50!).

(a) Write down the statistical hypothesis.

(b) Let the significance level be α, write down an equation to link the significance level and the
error probability that you want to control in the hypothesis testing.
(c) Suppose α = 0.05 and the p-value of your survey is 0.04, write down your conclusion and the
managerial implication formally.

2. Answer these true/false questions without providing any explanation.

(a) The significance level can be calculated based on a p-value.

(b) The p-value can be calculated based on a significance level.
(c) The critical values can be calculated based on a significance level.
(d) The critical values can be calculated based on a p-value.
(e) For a one-tailed test, once the p-value is lower than the significance level, we can reject H0
and conclude that, with the specified significance level, Ha is true.
(f) For a one-tailed test, once the p-value is higher than the significance level, we can accept H0
and conclude that, with the specified significance level, H0 is true.
(g) If H0 is true, Ha cannot be true.
(h) When we reject H0 , the significance level α is the probability for H0 to be true.
(i) The p-value is the probability of rejecting H0 when H0 is true.
(j) For a one-tailed test regarding the population mean, suppose the p-value is smaller than α.
In this case, observing a smaller p-value implies that the true population mean deviates more
from the assumed value.

3. According to the U.S. Bureau of Labor Statistics, the average weekly earnings of a production
worker in 1997 were $424.20. Suppose a labor researcher wants to test to determine whether this
figure is still accurate today. The researcher randomly selects 70 production workers from across
the United States and obtains a representative earnings statement for one week from each. The
resulting sample average is $432.70. Assuming a population standard deviation of $32, and a 5%
level of significance, determine whether the mean weekly earnings of a production worker have
changed.

4. Major cities around the world compete with each other in an effort to attract ne businesses. Some
of the criteria that businesses use to judge cities as potential locations for their headquarters
might include the labor pool; the environment, including work, governmental, and living; the tax
structure, the availability of skilled/educated labor, housing, education, medical care; and others.
Suppose in a study done several years ago, the city of Atlanta received a mean rating of 3.51
(on a scale of 1 to 5) on housing, but that since that time, considerable residential building has
occurred in the Atlanta area such that city leaders feel the mean might now be higher. They
hire a team of researchers to conduct a survey of businesses around the world to determine how

1
 businesses now rate the city on housing (and other variables). Sixty-four businesses take part in
the new survey, with a result that Atlanta receives a mean response of 3.71 on housing with a
sample standard deviation of 0.65. Assuming that such responses are normally distributed, use a
1% level of significance and these data to test to determine if the mean housing rating for the city
of Atlanta by businesses has significantly increased.
5. (15 points; modified from Problem 9.26) A study by Hewitt Associates showed that 76% of compa-
nies offer employees flexible scheduling. Suppose a researcher believes that in accounting firms this
percentage is lower. The researcher randomly selects 315 accounting firms and through interviews
determines that 223 of these firms have flexible scheduling. With a 1% level of significance, does
the test show enough evidence to conclude that a significantly lower proportion of accounting firms
offer employees flexible scheduling?
6. A manufacturing company produces bearings. One line of bearings is specified to be 2.65 centime-
ters (cm) in diameter. A major customer requires that the variance of the bearings be no more
than 0.001 cm2 . The producer is required to test the bearings before they are shipped, and so
the diameters of 20 bearings are measured with a precise instrument, recorded in the file ”bear-
ing.txt”. Assume bearing diameters are normally distributed. Use the data and α = 0.01 to test
to determine whether the variance is too high.