Exercise 

15.3

Self­Check Exercises
SC 15-1 Robin Zill and Stewart Griffiths own a small company that manufactures portable massage
tables in Hillsborough, North Carolina. Since they started the company, the number of tables
they have sold is represented by this time series:
Year 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996
Tables sold 42 50 61 75 92 111 120 127 140 138

(a) Find the linear equation that describes the trend in the number of tables sold by Robin
and Stewart.
(b) Estimate their sales of tables in 1998.
SC 15-2 The number of faculty-owned personal computers at the University of Ohio increased
dramatically between 1990 and 1995:
Year 1990 1991 1992 1993 1994 1995
Number of PCs 50 110 350 1,020 1,950 3,710

(a) Develop a linear estimating equation that best describes these data.
(b) Develop a second-degree estimating equation that best describes these data.
(c) Estimate the number of PCs that will be in use at the university in 1999, using both
equations.
(d) If there are 8,000 faculty members at the university, which equation is the better predictor?
Why?

Applications
15-12 Jeff Richardson invested his life savings and began a part-time carpet-cleaning business in
1986. Since 1986, Jeff’s reputation has spread and business has increased. The average
numbers of homes he has cleaned per month each year are:
Year 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996
Homes cleaned 6.4 11.3 14.7 18.4 19.6 25.7 32.5 48.7 55.4 75.7 94.3

(a) Find the linear equation that describes the trend in these data.
(b) Estimate the number of homes cleaned per month in 1997,1998, and 1999.
15-13 The owner of Progressive Builders is examining the number of solar homes started in the
region in each of the last 7 months:
Month June July Aug. Sept. Oct. Nov. Dec.
Number of homes 16 17 25 28 32 43 50

(a) Plot these data.

(b) Develop the linear estimating equation that best describes these data, and plot the line on
the graph from part (a) (let x units equal 1 month).
(c) Develop the second-degree estimating equation that best describes these data and plot
this curve on the graph from part (a).
(d) Estimate March sales using both curves you have plotted.
15-14 Richard Jackson developed an ergonomically superior computer mouse in 1989, and sales
have been increasing ever since. Data are presented below in terms of thousands of mice sold
per year.
Year 1989 1990 1991 1992 1993 1994 1995 1996
Number sold 82.4 125.7 276.9 342.5 543.6 691.5 782.4 889.5
 (a) Develop a linear estimating equation that best describes these data.
(b) Develop a second-degree estimating equation that best describes these data.
(c) Estimate the number of mice that will be sold in 1998, using both equations.
(d) If we assume the rate of increase in mouse sales will decrease soon based on supply and
demand, which model would be a better predictor for your answer in part (c)?
15-15 Mike Godfrey, the auditor of a state public school system, has reviewed the inventory records
to determine whether the current inventory holdings of textbooks are typical. The following
inventory amounts are from the previous 5 years:
Year 1991 1992 1993 1994 1995
Inventory ( $1,000) $4,620 $4,910 $5,490 $5,730 $5,990

(a) Find the linear equation that describes the trend in the inventory holdings.
(b) Estimate for him the value of the inventory for the year 1996.
15-16 The following table describes first-class postal rates from 1968 to 1996:
Year 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996
Rate (¢) 5 5 8 8 10 13 15 18 20 22 25 25 29 29 32

(a) Develop the linear estimating equation that best describes these data.
(b) Develop the second-degree estimating equation that best describes these data.
(c) Is there anything in the economic or political environment that would suggest that one or
the other of these two equations is likely to be the better predictor of postal rates?
15-17 Environtech Engineering, a company that specializes in the construction of antipollution
filtration devices, has recorded the following sales record over the last 9 years:
Year 1987 1988 1989 1990 1991 1992 1993 1994 1995
Sales ( $100,000) 13 15 19 21 27 35 47 49 57

(a) Plot these data.

(b) Develop the linear estimating equation that best describes these data, and plot this line on
the graph from part (a).
(c) Develop the second-degree estimating equation that best describes these data, and plot
this curve on the graph from part (a).
(d) Does the market to the best of your knowledge favor (b) or (c) as the more accurate
estimating method?
15-18 Here are data describing the air pollution rate (in ppm of particles in the air) in a western city:
Year 1980 1985 1990 1995
Pollution rate 220 350 800 2,450

(a) Would a linear or a second-degree estimating equation provide the better prediction of
future pollution in that city?
(b) Considering the economic, social, and political environment, would you change your
answer to part (a)?
(c) Describe how political and social action could change the effectiveness of either of the
estimating equations in part (a).
15-19 The State Department of Motor Vehicles is studying the number of traffic fatalities in the state
resulting from drunk driving for each of the last 9 years.
Year 1987 1988 1989 1990 1991 1992 1993 1994 1995
Deaths 175 190 185 195 180 200 185 190 205

(a) Find the linear equation that describes the trend in the number of traffic fatalities in the
state resulting from drunk driving.
(b) Estimate the number of traffic fatalities resulting from drunk driving that the state can
expect in 1996.
Exercise 15.4

Self­Check Exercise
SC 15-3 The Western Natural Gas Company has supplied 18, 20, 21, 25, and 26 billion cubic feet of
gas, respectively, for the years 1991 to 1995.
(a) Find the linear estimating equation that best describes these data.
(b) Calculate the percent of trend for these data.
(c) Calculate the relative cyclical residual for these data.
(d) In which years does the largest fluctuation from trend occur, and is it the same for both
methods?

Applications
15-20 Microprocessing, a computer firm specializing in software engineering, has compiled the
-following revenue records for the years 1989 to 1995
Year 1989 1990 1991 1992 1993 1994 1995
Revenue ( $100,000) 1.1 1.5 1.9 2.1 2.4 2.9 3.5

The second-degree equation that best describes the secular trend for these data is

Ŷ  2.119  0.375x  0.020x2, where 1992  0, and x units  1 year

(a) Calculate the percent of trend for these data.
(b) Calculate the relative cyclical residual for these data.
(c) Plot the percent of trend from part (a).
(d) In which year does the largest fluctuation from trend occur, and is it the same for both
methods?
15-21 The Bulls Eye department store has been expanding market share during the past 7 years,
posting the following gross sales in millions of dollars:
Year 1990 1991 1992 1993 1994 1995 1996
Sales 14.8 20.7 24.6 32.9 37.8 47.6 51.7

(a) Find the linear estimating equation that best describes the data.
(b) Calculate the percent of trend for these data.
(c) Calculate the relative cyclical residual for these data.
(d) In which years does the largest fluctuation from trend occur, and is it the same for both
methods?
15-22 Joe Honeg, the sales manager responsible for the appliance division of a large consumer-
products company, has collected the following data regarding unit sales for his division
during the last 5 years:
Year 1991 1992 1993 1994 1995
Units ( 10,000) 32 46 50 66 68

The equation describing the secular trend for appliance sales is

Yˆ  52.4  9.2 x, where 1993  0, and x units  1 year

(a) Calculate the percent of trend for these data.
(b) Calculate the relative cyclical residual for these data.
(c) Plot the percent of trend from part (a).
(d) In which year does the largest fluctuation from trend occur, and is it the same for both
 methods?
15-23 Suppose you are the capital budgeting officer of a small corporation whose financing
requirements over the last few years have been
Year 1989 1990 1991 1992 1993 1994 1995
Millions of dollars required 2.2 2.1 2.4 2.6 2.7 2.9 2.8

The trend equation that best describes these data is

Yˆ  2.53  0.13x, where 1992  0, and x units  1 year
(a) Calculate the percent of trend for these data.
(b) Calculate the relative cyclical residual for these data.
(c) In which year does the largest fluctuation from trend occur, and is it the same for both
methods?
(d) As the capital budgeting officer, what would this fluctuation mean for you and the
activities you perform?
15-24 Parallel Breakfast Foods has data on the number of boxes of cereal it has sold in each of the
last 7 years.
Year 1989 1990 1991 1992 1993 1994 1995
Boxes ( 10,000) 21.0 19.4 22.6 28.2 30.4 24.0 25.0

(a) Find the linear estimating equation that best describes these data.
(b) Calculate the percent of trend for these data.
(c) Calculate the relative cyclical residual for these data.
(d) In which year does the biggest fluctuation from the trend occur under each measure of
cyclical variation? Is this year the same for both measures? Explain.
15-25 Wombat Airlines, an Australian company, has gathered data on the number of passengers who
have flown on its planes during each of the last 5 years.
Year 1991 1992 1993 1994 1995
Passengers (in tens of thousands) 3.5 4.2 3.9 3.8 3.6

(a) Find the linear estimating equation that best describes these data.
(b) Calculate the percent of trend for these data.
(c) Calculate the relative cyclical residual for these data.
(d) Based on the data and your previous calculations, give a one-sentence summary of the
position in which Wombat Airlines finds itself.

Exercise 15.5

Self­Check Exercise
SC 15-4 Using the following percentages of actual to moving average describing the quarterly amount
of cash in circulation at the Village Bank in Carrboro, N.C. over a 4-year period, calculate the
seasonal index for each quarter.
Spring  Summer Fall Winter
1992 87 106 86 125
1993 85 110 83 127
1994 84 105 87 128
1995 88 104 88 124
Applications
15-26 The owner of The Pleasure-Glide Boat Company has compiled the following quarterly
figures regarding the company’s level of accounts receivable over the last 5 years ( $1,000):
Spring  Summer Fall Winter
1991 102 120 90 78
1992 110 126 95 83
1993 111 128 97 86
1994 115 135 103 91
1995 122 144 110 98

(a) Calculate a 4-quarter centered moving average.

(b) Find the percentage of actual to moving average for each period.
(c) Determine the modified seasonal indices and the seasonal indices.
15-27 Marie Wiggs, personnel director for a pharmaceutical company, recorded these percentage
absentee rates for each quarter over a 4-year period:
Spring Summer Fall Winter
1992 5.6 6.8 6.3 5.2
1993 5.7 6.7 6.4 5.4
1994 5.3 6.6 6.1 5.1
1995 5.4 6.9 6.2 5.3

(a) Construct a 4-quarter centered moving average and plot it on a graph along with the
-original data.
(b) What can you conclude about absenteeism from part (a)?
15-28 Using the following percentages of actual to moving average describing the seasonal sales of
sporting goods over a 5-year period, calculate the seasonal index for each season.
Year  Baseball Football Basketball  Hockey
1992 96 128 116 77
1993 92 131 125 69
1994 84 113 117 84
1995 97 118 126 89
1996 91 121 124 81

15-29 A large manufacturer of automobile springs has determined the following percentages of
actual to moving average describing the firm’s quarterly cash needs for the last ,6 years:
Spring  Summer  Fall  Winter
1990 108 128 94 70
1991 112 132 88 68
1992 109 134 84 73
1993 110 131 90 69
1994 108 135 89 68
1995 106 129 93 72

Calculate the seasonal index for each quarter. Comment on how it compares to the indices
you calculated for Exercise 15-26.
15-30 A university’s dean of admissions has compiled the following quarterly enrollment figures for
the previous 5 years (100):
Fall  Winter Spring  Summer 
1991 220 203 193 84
1992 235 208 206 76
1993 236 206 209 73
 1994 241 215 206 92
1995 239 221 213 115

(a) Calculate a 4-quarter centered moving average.

(b) Find the percentage of actual to moving average for each period.
(c) Determine the modified seasonal indices and the seasonal indices.
15-31 The Ski and Putt Resort, a combination of ski slopes and golf courses, has just recently
tabulated its data on the number of customers (in thousands) it has had during each season of
the last 5 years. Calculate the seasonal index for each quarter. If 15 people are employed in the
summer, what should winter employment be, assuming both sports have equal labor
requirements?
Spring  Summer  Fall  Winter
1991 200 300 125 325
1992 175 250 150 375
1993 225 300 200 450
1994 200 350 225 375
1995 175 300 200 350

15-32 David Curl Builders has collected quarterly data on the number of homes it has started during
the last 5 years.
Spring  Summer  Fall  Winter
1991 8 10 7 5
1992 9 10 7 6
1993 10 11 7 6
1994 10 12 8 7
1995 11 13 9 8

(a) Calculate the seasonal index for each quarter.

(b) If David’s working capital needs are related directly to the number of starts, by how
much should his working capital need decrease between summer and winter?

Exercise 15.7

Self­Check Exercise
SC 15-5 A state commission designed to monitor energy consumption assembled the following
-seasonal data regarding natural gas consumption, in millions of cubic feet:
Winter Spring Summer Fall
1992 293 246 231 282
1993 301 252 227 291
1994 304 259 239 296
1995 306 265 240 300

(a) Determine the seasonal indices and deseasonalize these data (using a 4-quarter centered
moving average).
(b) Calculate the least-squares line that best describes these data.
(c) Identify the cyclical variation in these data by the relative cyclical residual method.
 (d) Plot the original data, the deseasonalized data, and the trend.

Applications
15-37 An environmental agency has been watching New York air quality over a 5-year period and
has assembled the following seasonal data regarding amount of contaminants (in parts per
million) in the air:
Year Winter Spring Summer Fall
1992 452 385 330 385
1993 474 397 356 399
1994 494 409 375 415
1995 506 429 398 437
1996 527 454 421 482

(a) Determine the seasonal indices and deseasonalize these data (using a 4-quarter centered
moving average).
(b) Calculate the least-squares line that best describes these data.
(c) Identify the cyclic variation in these data by the relative cyclical residual method.
(d) Plot the original data, the deseasonalized data, and the trend.
15-38 The following data describe the marketing performance of a regional beer producer:
Sales by Quarter ( $100,000)

Year I II III IV
1991 19 24 38 25
1992 21 28 44 23
1993 23 31 41 23
1994 24 35 48 21

(a) Calculate the seasonal indices for these data. (Use a 4-quarter centered moving
average.)
(b) Deseasonalize these data using the indices from part (a).
15-39 For Exercise 15-38:
(a) Find the least-squares line that best describes the trend in deseasonalized beer sales.
(b) Identify the cyclical component in this time series by computing the percent
of trend.

Chapter Concepts Test

   Circle the correct answer or fill in the blank. Answers are in the back of the book.
TF 1. Time-series analysis is used to detect patterns of change in statistical information
over regular intervals of time.
TF 2. Secular trends represent the long-term direction of a time series.
TF 3 ‘When coding time values, we subtract from each value the smallest time value in
the series; hence, the code of the smallest value is zero.
 TF 4. When using the least-squares method to determine a second-degree equation of best
fit, the values of four numerical constants must be determined.
TF 5. Time-series analysis helps us to analyze past trends, but it cannot aid us in future
uncertainties.
TF 6. When we are predicting far into the future, a second-degree equation usually gives
more accurate predictions than a linear equation.
TF 7. When using the residual method, we assume that the cyclical component explains
most of the variation left unexplained by the trend component.
TF 8. The relative cyclical residual can be computed for an entry in a time series by
subtracting 10 from the percent of trend for that entry.
TF 9. The repetitive movement around a trend line in a 2-year period is best described as a
seasonal variation.
TF 10. Once seasonal indices are computed for a time series, the series can be
deseasonalized so that only the trend component remains.
TF 11. The percent of trend should not be used for predicting future cyclical variations.
TF 12. Over time, random movements tend to counteract one another in irregular variation
in a time series.
TF 13. Before the percent of trend can be calculated, a trend line (graph of must first be
calculated.
TF 14. If a time series contains an odd number of elements, then the coding for some of the
entries will be in half-units.
TF 15. To be considered a time series, a group of statistical information must have been
accumulated at regular intervals.
TF 16. Of the four types of variation, cyclical is the most difficult to predict.
TF 17. Seasonal variation is a repetitive and predictable variation around the trend line
within a year.
TF 18. To picture the isolated cyclical component of a time series, we may graph either the
percent of. trend values or the relative cyclical residual values.
TF 19. The adjusted seasonal indices must always sum to 400.
TF 20. A time series should be deseasonalized after the trend or cyclical components of the
time series have been identified.
TF 21. Removing the highest and lowest actual-to-moving-average values from each period
when computing seasonal indices reduces the extreme cyclical and irregular
variations.
ABCDE 22. A time series of annual data can contain which of the following components?
(a) Secular trend.
(b) Cyclical fluctuation.
(c) Seasonal variation.
(d) All of these.
(e) (a) and (b) but not (c).
ABCD 23. Suppose you were considering a time series of data for the quarters of 1992 and
1993. The third quarter of 1993 would be coded as:
(a) 2.
(b) 3.
(c) 5.
(d) 6.
ABCD 24. Suppose that a particular time series should be fitted with a parabolic curve. The
general form for this second-degree equation is  = a + bx + cx2. What do the x’s
represent in this formula?
(a) Coded values of the time variables.
(b) A numerical constant that is determined by a formula.
(c) Estimates of the dependent variable.
(d) None of these.
ABCDE 25. Assume that a time series with annual data for the years 1988-1996 is described well
by the second-degree equation = 5 + 3x + 9x2. Based only on this secular trend, what
is the forecast value for 1997?
(a) 161.
(b) 245.
(c) 347.
(d) 293.75.
(e) 200.75.
ABCD 26. Suppose that  = 10 + 3x describes well an annual time series for 1987-1993. If the
actual value of Y for 1990 is 8, what is the percent of trend for 1990?
(a) 125%.
(b) 112.5%.
(c) 90%.
(d) 80%.
ABCD 27. A time series for the years 1985-1996 had the following relative cyclical residuals,
in chronological order: –1%, –2%, 1%, 2%, –1%, –2%, 1%, 2%, –1%, –2%, 1%,
2%. The relative cyclical residual for 1997 should be:
(a) 3%.
(b) –1%.
(c) –2%.
(d) Cannot be determined from information given.
ABCD 28. Assume that you have been given quarterly sales data for a 5-year period. To use the
ratio-to-moving-average method of computing a seasonal index, your first step
would be:
(a) Compute the 4-quarter moving average.
(b) Discard highest and lowest values for each quarter.
(c) Calculate the 4-quarter moving total.
 (d) None of these.
Questions 29 through 31 deal with a seasonal index being computed, using the ratio-
to-moving-average method for quarterly data from 1992-1996. The percentages of
actual to moving average for the third quarter of each year are 1992,109.0;
1993,112.8; 1994, 110.0; 1995,108.0; 1996, 104.6.
ABCDE 29. What is the unadjusted index for the third quarter?
(a) 108.88.
(b) 109.0.
(c) 110.23.
(d) 110.96.
(e) None of these.
ABCDE 30. Assume that the total of the unadjusted indices for the four quarters is 404.04. If the
unadjusted index for the first quarter is 97.0, what is the adjusted seasonal index for
the first quarter?
(a) 96.03.
(b) 97.98.
(c) 24.01.
(d) 99.00.
(e) Cannot be determined from the information given.
ABCD 31. The adjusted seasonal index for the fourth quarter is 95.0. If the deseasonalized
trend line that was calculated to estimate quarterly sales is   = 400 + 9x, what would
be the seasonalized sales estimate for the fourth quarter of 1994?
(a) 499.7.
(b) 643.0.
(c) 610.85.
(d) 676.8.
ABCDE 32. If a time series has an even number of years, and we use coding, then each coded
interval is equal to:
(a) 1 year.
(b) 2 years.
(c) 1 month.
(d) 6 months.
(e) None of these.
A B C D E F 33. A method used to deal with cyclical variation when the cyclical component does not
explain most of the variation left unexplained by the trend component is:
(a) Spearman analysis.
(b) Specific analysis.
(c) Second-degree analysis.
(d) Relative cyclical residual.
(e) All of these.
 (f) None of these.
A B C D E 34. For a given year, if an adjusted seasonal index for some period is greater than 100,
then the following must be true:
(a) The adjusted index for some other period is > 100.
(b) The adjusted index for some other period is < 100.
(c) The adjusted index for some other period is = 100.
(d) (a) and (b) but not (c).
(e) None of the above.
A B C D 35. If the percent of trend for a particular year in a time series is greater than 100%, then
for this year:
(a) The actual time-series value lies below the trend line and the relative cyclical
residual is positive.
(b) The actual time-series value lies below the trend line and the relative cyclical
residual is positive.
(c) The actual time-series value lies above the trend line and the relative cyclical
residual is negative.
(d) The actual time-series value lies above the trend line and the relative cyclical
residual is positive.
A B C D E 36. Which of the following are common reasons for studying both secular trends and
seasonal variation:
(a) To allow the elimination of the component from the series.
(b) To describe past patterns.
(c) To project past patterns into the future.
(d) All of the above.
(e) None of the above
37. Dividing each actual value in a time series by the corresponding trend value, and
multiplying by 100, gives the ____________________.
38. Repetitive and predictable movement around the trend line in 1 year or less is
____________________variation.
39. ____________________ variation in a time series is characterized by un-
predictable, random movement and usually occurs over short intervals.
40. ____________________ variation is the time-series component that oscillates above
and below the trend line for periods longer than one year.
41. Using seasonal indices to remove effects of seasonality from a time series is known
as ____________________ the time series.
42. The first step in computing the seasonal indices for some time-series data is to
calculate the ____________________ .
43. The steady increase in the cost of living recorded by the CPI (Consumer Price
Index) is an example of a ____________________.
 44. The result of discarding the highest and lowest values before
averaging is called a __________________.
45. A ____________________ is a projection of past patterns
into the future.