ASBE 5e Solutions for Instructors
Exam Review Questions
Chapters 1-4
1.
a. This is an inferential statistic. We are inferring the true default rate from a sample.
b. This is a descriptive statistic. We are simply describing the percentage from our class
who use Verizon.
c. This is an inferential statistic. We are inferring the true average battery life from a
sample.
2.
a. Ethical.
b. Ethical.
c. Not ethical. A statistician should not sway his or her results based on the client’s wishes.
3.
This best illustrates conclusion from a small sample. A sample of one person does not
prove that driving without a seat belt is not risky. The risk is determined by looking at
samples of drivers who use their seatbelts and samples of drivers who don’t use their
seatbelts and comparing their injury rates.
4.
a. Numerical
b. Categorical
c. Numerical
5.
a. Ratio. There is a defined zero point.
b. Ordinal. There is ranking implied but the intervals between rankings cannot be assumed
equal.
c. Nominal. The type of charge card is simply a category.
6.
a. Continuous
b. Continuous
c. Discrete
7.
a. Convenience
b. Simple random
c. Systematic
8.
a. Sampling is the only choice. Destructive testing prohibits the use of a census.
b. Sampling makes sense. It would not be wise to deplete resources in order to take a
census.
c. The inherent bias in Excel’s random number generator might lead one to consider a
census because it is possible that a true random sample might not be possible if the
sample items are selected by using Excel.
ASBE 5e Solutions for Instructors
9.
a. Correct. We can treat Likert data as interval if we assume the distances between scale
points are equal.
b. Not correct. Cross-sectional data are taken at a point in time, not over several time
intervals.
c. Not correct. There are circumstances that prevent us from taking a census and there is the
possibility that a census is flawed resulting in undercounting certain parts of the
population.
10.
a. False. Sampling error can only be reduced (not eliminated) by increasing the sample size.
b. True. Selection bias refers to people self-selecting to be in a sample which can result in a
sample that does not represent the population.
c. True. In order to match the random numbers generated for the sample with an item in the
population, one must have a list of the population with numbers assigned to each
member.
11.
a.
Histogram of Rider Ages
60
50
Percent
40
30
20
10
0
Midpoint of Bin
b. Sturges’ Rule says to use approximately 6 bins. MegaStat created a histogram with 7
bins so that the bin width was 10 years.
c. The distribution is strongly skewed right.
d.
ASBE 5e Solutions for Instructors
DotPlot
0
10
20
30
40
Rider Ages
50
60
70
e. The displays look similar because both are skewed right. But the dot plot shows that
there is a gap in ages from about 25-35.
12.
a. False. Histograms show the distribution of one variable only. Correlation describes the
relationship between two variables.
b. False. Pyramid charts make it difficult to interpret. Column or bar charts with rectangular
shapes are easier to interpret.
c. True. Correlation measures the strength of the linear relationship between two variables.
If two variables have an inverse relationship their correlation coefficient will be
negative.
13.
a. A pie chart would work because the three party designations make up a whole and we
would be comparing percentages or counts.
b. A pie chart would not work because the retail prices don’t make up counts or
percentages of a whole.
c. A pie chart would not work because the labor costs for each automaker do not tally
counts or percentages of a whole.
14.
x  12, s  5.7, C.V .  47.5%
15.
a. x  59.3 , median = 58.5, and mode = 62
b. The mean and median are appropriate for numerical, continuous data such as ages. The
mode is more appropriate for categorical data or integer valued numerical data with a
small range. For this data set choose either the mean or median because the data are not
highly skewed.
c. Q1 = 55, Q 2 = 62.25. 25% of CEO’s are 55 years or younger. 75% of CEOs are 62.25
years or younger.
d. The boxplot shows a fairly symmetric distribution with one high outlier.
BoxPlot
40
50
60
70
CEO Ages
80
90
ASBE 5e Solutions for Instructors
16.
a.
180
160
Salary
140
120
100
80
60
0
5
10
Years
15
20
There is a weak positive relationship between years of service and salary.
b. r = .5656. The correlation coefficient is not strong but is surprisingly higher than the
scatterplot indicates it should be.
17.
a. False. The median is less than the mean when the distribution is skewed right.
b. True. The geometric mean can only be calculated for positive values.
c. False. The midrange uses only the low and high value of the data set and is therefore not
resistant to outliers.
18.
a. False. z = (81-52)/15 = 1.93. This would not be considered an outlier because the z-score
is less than 2.
b. True. The empirical rule states that approximately 68% of observations are within one
standard deviation of the mean.
c. True. The C.V. = σ/µ. 128/640 = .2 or 20%.
19.
a. True. The log scale brings the magnitudes closer making it easier to compare.
b. False. Log scales are more difficult to interpret.
c. True.