Appendix D Self-Test Solutions and Answers to Even-Numbered Exercises 1051

12. a. 2428, 2490, 2451 16. I  83

Manufacturing shipments increased slightly in con-
stant dollars
18. a. 151, 197, 143, 178
b. 3043, 3132, 3050
b. I  170
c. PPI
300(18.00)  400(4.90)  850(15.00) 20. IJan  73.5, IMar  70.1
14. I  (100)
350(18.00)  220(4.90)  730(15.00)
20,110 22. I  182.7
 (100)  110
18,328
95(1200)  75(1800)  50(2000)  70(1500) 24. $36,082; $32,528; $27,913; $34,387; $40,551; $42,651;
15. I  (100)
120(1200)  86(1800)  35(2000)  60(1500) $46,458; $56,324
 99
Quantities are down slightly 26. I  143; quantity is up 43%

Chapter 18
1. The following table shows the calculations for parts (a), (b), and (c):

Time Absolute Value Squared Absolute Value

Series Forecast of Forecast Forecast Percentage of Percentage
Week Value Forecast Error Error Error Error Error
1 18
2 13 18 5 5 25 38.46 38.46
3 16 13 3 3 9 18.75 18.75
4 11 16 5 5 25 45.45 45.45
5 17 11 6 6 36 35.29 35.29
6 14 17 3 3 9 21.43 21.43
Totals 22 104 51.30 159.38

22
a. MAE   4.4
5
104
b. MSE   20.8
5
159.38
c. MAPE   31.88
5
d. Forecast for week 7 is 14

2. The following table shows the calculations for parts (a), (b), and (c):

Time Absolute Value Squared Absolute Value

Series Forecast of Forecast Forecast Percentage of Percentage
Week Value Forecast Error Error Error Error Error
1 18
2 13 18.00 5.00 5.00 25.00 38.46 38.46
3 16 15.50 0.50 0.50 0.25 3.13 3.13
4 11 15.67 4.67 4.67 21.81 42.45 42.45
5 17 14.50 2.50 2.50 6.25 14.71 14.71
6 14 15.00 1.00 1.00 1.00 7.14 7.14
Totals 13.67 54.31 70.21 105.86
1052 Appendix D Self-Test Solutions and Answers to Even-Numbered Exercises

13.67 d. The three-week moving average provides a better fore-

a. MAE   2.73
5 cast since it has a smaller MSE
54.31 e. Smoothing constant  .4
b. MSE   10.86
5
105.89 Time Squared
c. MAPE   21.18
5 Series Forecast Forecast
Week Value Forecast Error Error
d. Forecast for week 7 is
18  13  16  11  17  14 1 18
 14.83 2 13 18.00 5.00 25.00
6 3 16 16.00 0.00 0.00
4 11 16.00 5.00 25.00
363 5 17 14.00 3.00 9.00
4. a. MSE   60.5
6 6 14 15.20 1.20 1.44
Forecast for month 8 is 15 Total 60.44
216.72 60.44
b. MSE   36.12 MSE   12.09
6 5
Forecast for month 8 is 18
c. The average of all the previous values is better because
MSE is smaller The exponential smoothing forecast using α  .4 provides a
better forecast than the exponential smoothing forecast us-
5. a. The data appear to follow a horizontal pattern ing α  .2 since it has a smaller MSE
b. Three-week moving average. 6. a. The data appear to follow a horizontal pattern
110
b. MSE   27.5
Time Squared 4
Series Forecast Forecast The forecast for week 8 is 19
Week Value Forecast Error Error 252.87
c. MSE   42.15
1 18 6
2 13 The forecast for week 7 is 19.12
3 16 d. The three-week moving average provides a better fore-
4 11 15.67 4.67 21.78 cast since it has a smaller MSE
5 17 13.33 3.67 13.44 e. MSE  39.79
6 14 14.67 0.67 0.44
The exponential smoothing forecast using α  .4 pro-
Total 35.67 vides a better forecast than the exponential smoothing
35.67 forecast using α  .2 since it has a smaller MSE
MSE   11.89
3
8. a.

(11  17  14) Week 4 5 6 7 8 9 10 11 12

The forecast for week 7   14 Forecast 19.33 21.33 19.83 17.83 18.33 18.33 20.33 20.33 17.83
3
c. Smoothing constant  .2
b. MSE  11.49
Prefer the unweighted moving average here; it has a
Time Squared
Series Forecast Forecast smaller MSE
Week Value Forecast Error Error c. You could always find a weighted moving average at
least as good as the unweighted one; actually the un-
1 18
2 13 18.00 5.00 25.00 weighted moving average is a special case of the
3 16 17.00 1.00 1.00 weighted ones where the weights are equal
4 11 16.80 5.80 33.64 10. b. The more recent data receive the greater weight or im-
5 17 15.64 1.36 1.85
portance in determining the forecast; the moving aver-
6 14 15.91 1.91 3.66
ages method weights the last n data values equally in
Total 65.15 determining the forecast
65.15
MSE   13.03 12. a. The data appear to follow a horizontal pattern
5
b. MSE(3-Month)  .12
MSE(4-Month)  .14
The forecast for week 7 is .2(14)  (1  .2)15.91 Use 3-Month moving averages
 15.53 c. 9.63
 Appendix D Self-Test Solutions and Answers to Even-Numbered Exercises 1053

13. a. The data appear to follow a horizontal pattern b. The methods discussed in this section are only
b. applicable for a time series that has a horizontal pat-
tern, so if there is a really a long-term linear trend in
the data, the methods discussed in this section are not
3-
Month appropriate
Time- Moving c. The time series plot for the data for years 2002–2008
Series Average α ⴝ .2 exhibits a horizontal pattern; it seems reasonable to con-
Month Value Forecast (Error)2 Forecast (Error)2 clude that the extreme values observed in 1997 and
1 240 2001 are more attributable to viewer interest in the per-
2 350 240.00 12100.00 formance of Tiger Woods; basing the forecast on years
3 230 262.00 1024.00
2002–2008 does seem reasonable, but, because of the
4 260 273.33 177.69 255.60 19.36
5 280 280.00 0.00 256.48 553.19 injury that Tiger Woods experienced in the 2008 reason,
6 320 256.67 4010.69 261.18 3459.79 if he is able to play in the 2009 Masters, then the rating
7 220 286.67 4444.89 272.95 2803.70 for 2009 may be significantly higher than suggested by
8 310 273.33 1344.69 262.36 2269.57 the data for years 2002–2008
9 240 283.33 1877.49 271.89 1016.97
10 310 256.67 2844.09 265.51 1979.36 17. a. The time series plot shows a linear trend
11 240 286.67 2178.09 274.41 1184.05 n n
12 230 263.33 1110.89 267.53 1408.50
兺t 15 兺Y t
55
Totals 17,988.52 27,818.49 b. t  t1
 3 Y t1
  11
n 5 n 5
MSE (3-Month)  17,988.52/9  1998.72
MSE (α  .2)  27,818.49/11  2528.95 兺(t  t̄ )(Yt  Ȳ )  21 兺(t  t̄ )2  10
n

Based on the preceding MSE values, the 3-Month 兺(t  t)(Y  Y)

t
21
moving averages appear better; however, exponential b1  t1
  2.1
n
10
smoothing was penalized by including month 2, which 兺(t  t) 2

was difficult for any method to forecast; using only the t1

errors for months 4 to 12, the MSE for exponential b0  Ȳ  b1t̄  11  (2.1)(3)  4.7
smoothing is
Tt  4.7  2.1t
MSE(α  .2)  14,694.49/9  1632.72
c. T6  4.7  2.1(6)  17.3
Thus, exponential smoothing was better considering
months 4 to 12 18. Forecast for week 6 is 21.16
c. Using exponential smoothing, 20. a. The time series plot exhibits a curvilinear trend
F13  αY12  (1  α)F12 b. Tt  107.857  28.9881t  2.65476t2
c. 45.86
.20(230)  .80(267.53)  260
21. a. The time series plot shows a linear trend
14. a. The data appear to follow a horizontal pattern n n
b. Values for months 2–12 are as follows: 兺t 45 兺Y t
108
105.00 114.00 115.80 112.56 105.79 110.05 b. t  t1
 5 Y t1
  12
n 9 n 9
120.54 126.38 118.46 106.92 104.85
兺(t  t̄ )(Yt  Ȳ )  87.4 兺(t  t̄ )2  60
MSE  510.29
n
c. Values for months 2–12 are as follows:
兺(t  t)(Y  Y)
t
87.4
105.00 120.00 120.00 112.50 101.25 110.63 b1  t1
  1.4567
n
60
127.81 133.91 116.95 98.48 99.24 兺(t  t)
t1
2

MSE  540.55
b0  Ȳ  b1t̄  12  (1.4567)(5)  4.7165
Conclusion: A smoothing constant of .3 is better than a
smoothing constant of .5 since the MSE is less for 0.3 Tt  4.7165  1.4567t
16. a. The time series plot indicates a possible linear trend in c. T10  4.7165  1.4567(10)  19.28
the data; this could be due to decreasing viewer interest 22. a. The time series plot shows a downward linear trend
in watching the Masters, but closer inspection of the b. Tt  13.8  .7t
data indicates that the two highest ratings correspond to c. 8.2
years 1997 and 2001, years in which Tiger Woods won d. If SCF can continue to decrease the percentage
the tournament; the pattern observed may be simply due of funds spent on administrative and fund-raising by
to the effect Tiger Woods has on ratings and not neces- .7% per year, the forecast of expenses for 2015 is
sarily on any long-term decrease in viewer interest 4.70%
1054 Appendix D Self-Test Solutions and Answers to Even-Numbered Exercises

24. a. The time series plot shows a linear trend b. A portion of the Minitab regression output follows:
b. Tt  7.5623  .07541t
c. 6.7328 The regression equation is
d. Given the uncertainty in global market conditions, Revenue = 70.0 + 10.0 Qtr1 + 105
making a prediction for December using only time is Qtr2 + 245 Qtr3
not recommended
26. a. A linear trend is not appropriate Quarter 1 forecast is 80
b. Tt  5.702  2.889t  1618t2 Quarter 2 forecast is 175
c. 17.90 Quarter 3 forecast is 315
28. a. The time series plot shows a horizontal pattern, but Quarter 4 forecast is 70
there is a seasonal pattern in the data; for instance, in c. A portion of the Minitab regression output follows
each year the lowest value occurs in quarter 2 and the
highest value occurs in quarter 4 The regression equation is
b. A portion of the Minitab regression output is shown; Revenue = -70.1 + 45.0 Qtr1 + 128
Qtr2 + 257 Qtr3 + 11.7 Period
The regression equation is
Value = 77.0 - 10.0 Qtr1 - 30.0
Quarter 1 forecast  is 221
Qtr2 - 20.0 Qtr3
Quarter 2 forecast  is 315
Quarter 3 forecast  is 456
c. The quarterly forecasts for next year are as follows: Quarter 4 forecast  is 211
Quarter 1 forecast  77.0  10.0(1)  30.0(0)
34. a. The time series plot shows seasonal and linear trend
 20.0(0)  67
effects
Quarter 2 forecast  77.0  10.0(0)  30.0(1)
b. Note: Jan  1 if January, 0 otherwise; Feb  1 if
 20.0(0)  47
February, 0 otherwise; and so on
Quarter 3 forecast  77.0  10.0(0)  30.0(0)
 20.0(1)  57 A portion of the Minitab regression output follows:
Quarter 4 forecast  77.0  10.0(0)  30.0(0)
 20.0(0)  77 The regression equation is
30. a. There appears to be a seasonal pattern in the data and Expense = 175 - 18.4 Jan - 3.72 Feb +
perhaps a moderate upward linear trend 12.7 Mar + 45.7 Apr + 57.1
b. A portion of the Minitab regression output follows: May + 135 Jun + 181 Jul + 105
Aug + 47.6 Sep + 50.6 Oct +
The regression equation is 35.3 Nov + 1.96 Period
Value = 2492 - 712 Qtr1 - 1512
Qtr2 + 327 Qtr3 c. Note: The next time period in the time series is
Period  37 (January of Year 4); the forecasts for Janu-
c. The quarterly forecasts for next year are as follows: ary–December are 229; 246; 264; 299; 312; 392; 440;
Quarter 1 forecast is 1780 366; 311; 316; 302; 269
Quarter 2 forecast is 980 35. a. The time series plot indicates a linear trend and a sea-
Quarter 3 forecast is 2819 sonal pattern
Quarter 4 forecast is 2492 b.
d. A portion of the Minitab regression output follows:
Time Four-Quarter Centered
The regression equation is Series Moving Moving
Value = 2307 - 642 Qtr1 - 1465 Year Quarter Value Average Average
Qtr2 + 350 Qtr3 + 23.1 t 1 1 4

2 2
The quarterly forecasts for next year are as follows:
3.50
Quarter 1 forecast is 2058 3 3 3.750
Quarter 2 forecast is 1258 4.00
Quarter 3 forecast is 3096 4 5 4.125
Quarter 4 forecast is 2769 4.25
2 1 6 4.500
32. a. The time series plot shows both a linear trend and sea- 4.75
sonal effects
 Appendix D Self-Test Solutions and Answers to Even-Numbered Exercises 1055

Time Four-Quarter Centered Deseasonalized

Series Moving Moving Year Quarter Value
Year Quarter Value Average Average
2 1 4.979
2 3 5.000 2 4.021
5.25 3 5.834
3 5 5.375 4 5.877
5.50 3 1 5.809
4 7 5.875 2 8.043
6.25 3 7.001
3 1 7 6.375 4 6.717
6.50
2 6 6.625
6.75 b. Let Period  1 denote the time series value in Year 1—
3 6 Quarter 1; Period  2 denote the time series value in
Year 1—Quarter 2; and so on; a portion of the Minitab
4 8 regression output treating Period as the independent
variable and the Deseasonlized Values as the values of
the dependent variable follows:
c.
The regression equation is
Time Centered Seasonal- Deseasonalized Value = 2.42 + 0.422
Series Moving Irregular Period
Year Quarter Value Average Component
1 1 4
2 2 c. The quarterly deseasonalized trend forecasts for Year 4
3 3 3.750 0.800 (Periods 13, 14, 15, and 16) are as follows:
4 5 4.125 1.212 Forecast for quarter 1 is 7.906
2 1 6 4.500 1.333 Forecast for quarter 2 is 8.328
2 3 5.000 0.600 Forecast for quarter 3 is 8.750
3 5 5.375 0.930
Forecast for quarter 4 is 9.172
4 7 5.875 1.191
3 1 7 6.375 1.098 d. Adjusting the quarterly deseasonalized trend forecasts
2 6 6.625 0.906 provides the following quarterly estimates:
3 6 Forecast for quarter 1 is 9.527
4 8
Forecast for quarter 2 is 6.213
Forecast for quarter 3 is 7.499
Forecast for quarter 4 is 10.924
Seasonal- Adjusted 38. a. The time series plot shows a linear trend and seasonal
Irregular Seasonal Seasonal effects
Quarter Values Index Index
b. 0.71 0.78 0.83 0.97 1.02 1.30 1.50 1.23
1 1.333 1.098 1.216 1.205 0.98 0.99 0.93 0.79
2 0.600 0.906 0.752 0.746 c.
3 0.800 0.930 0.865 0.857
4 1.212 1.191 1.201 1.191
Deaseasonalized
Total 4.036
Month Expense
Adjustment for 4.000
  0.991 1 239.44
seasonal index 4.036 2 230.77
3 246.99
4 237.11
36. a.
5 235.29
6 242.31
Deseasonalized 7 240.00
Year Quarter Value 8 235.77
9 244.90
1 1 3.320 10 242.42
2 2.681 11 247.31
3 3.501
4 4.198 (Continued)
1056 Appendix D Self-Test Solutions and Answers to Even-Numbered Exercises

time period, and then decreases again. There may also

Deaseasonalized be some linear trend in the data
Month Expense b.
12 246.84
13 253.52 Time Period Adjusted Seasonal Index
14 262.82 12–4 A.M. 0.3256
15 259.04 4–8 A.M. 0.4476
16 252.58 8–12 noon 1.3622
17 259.80 12–4 P.M. 1.6959
18 253.85 4–8 P.M. 1.4578
19 266.67 8–12 midnight 0.7109
20 272.36
21 265.31
22 272.73 c. The following Minitab output shows the results of fitting
23 274.19 a linear trend equation to the deseasonalized time series:
24 278.48
25 274.65
26 269.23 The regression equation is
27 277.11 Deseasonalized Power = 63108 + 1854 t
28 288.66
29 284.31 Deaseasonalized Power (t  19)  63,108  1854(19)
30 300.00
31 280.00  98,334
32 268.29 Forecast for 12–4 P.M.  1.6959(98,334)  166,764.63 or
33 295.92 approximately 166,765 kWh
34 297.98
Deaseasonalized Power (t  20)  63,108  1854(20)
35 301.08
36 316.46  100,188
Forecast for 4–8 P.M.  1.4578(100,188)  146,054.07 or
approximately 146,054 kWh
d. Let Period  1 denote the time series value in January Thus, the forecast of power consumption from noon to
-Year 1; Period  2 denote the time series value in Feb-
8 P.M. is 166,765  146,054  312,819 kWh
ruary-Year 2; and so on; a portion of the Minitab re-
gression output treating Period as the independent 42. a. The time series plot indicates a horizontal pattern
variable and the Deseasonlized Values as the values of b. MSE(α  .2)  1.40
the dependent variable follows: MSE(α  .3)  1.27
MSE(α  .4)  1.23
The regression equation is A smoothing constant of α  .4 provides the best forecast
Deseasonalized Expense = 228 + 1.96 because it has a smaller MSE
Period c. 31.00
44. a. There appears to be an increasing trend in the data
e. b. A portion the Minitab regression output follows (Note:
t  1 corresponds to 2001, t  2 corresponds to 2002,
Month Monthly Forecast
and so on)
January 213.37
February 235.93 The regression equation is
March 252.69 Balance($) = 1984 + 146 t
April 297.21
May 314.53 The forecast for 2009 (t  9) is Balance($)
June 403.42
 1984  146(9)  $3298
July 486.42
August 386.52 c. A portion of the Minitab regression output follows
September 309.88 (Note: t  1 corresponds to 2001, t  2 corresponds to
October 314.98 2002, and so on)
November 297.71
December 254.44 The regression equation is
Balance($) = 2924 - 419 t + 62.7 tsq
40. a. The time series plot indicates a seasonal effect; power
consumption is lowest in the time period 12–4 A.M., The forecast for 2009 (t  9) is Balance ($)  2924 
steadily increases to the highest value in the 12–4 P.M. 419(9)  62.7(9)2  $4232
 Appendix D Self-Test Solutions and Answers to Even-Numbered Exercises 1057

d. The quadratic trend equation provides the best forecast d. Hudson Marine experiences the largest seasonal in-
accuracy for the historical data crease in quarter 2; since this quarter occurs prior to the
e. Linear trend equation peak summer boating season, this result seems reason-
46. a. The forecast for July is 236.97 able, but the largest seasonal effect is the seasonal de-
Forecast for August, using forecast for July as the ac- crease in quarter 4; this is also reasonable because of
tual sales in July, is 236.97 decreased boating in the fall and winter
Exponential smoothing provides the same forecast for
every period in the future; this is why it is not usually
recommended for long-term forecasting Chapter 19
b. Using Minitab’s regression procedure we obtained the 1. n  27 cases with a value different than 150
linear trend equation Normal approximation μ  .5n  .5(27)  13.5
Tt  149.72  18.451t σ  兹.25 n  兹.25(27)  2.5981
Forecast for July is 278.88 With the number of plus signs  22 in the upper tail,
Forecast for August is 297.33 use continuity correction factor as follows
c. The proposed settlement is not fair since it does not ac- 21.5  13.5
count for the upward trend in sales; based upon trend
projection, the settlement should be based on forecasted
P(x
21.5)  P z
冢 2.5981 冣 P(z
3.08)

lost sales of $278,880 in July and $297,330 in August p-value  (1.0000  .9990)  .0010
p-value  .01; reject H0; conclude population median  150
48. a. The time series plot shows a linear trend
b. Tt  5  15t 2. Dropping the no preference, the binomial probabilities
The slope of 15 indicates that the average increase in for n  9 and p  .50 are as follows
sales is 15 pianos per year
c. 85, 100 x Probability x Probability
50. a. 0 0.0020 5 0.2461
1 0.0176 6 0.1641
2 0.0703 7 0.0703
Quarter Adjusted Seasonal Index
3 0.1641 8 0.0176
1 1.2717 4 0.2461 9 0.0020
2 0.6120
3 0.4978 Number of plus signs  7
4 1.6185
P(x
7)  P(7)  P(8)  P(9)
 .0703  .0176  .0020
4  .0899
Note: Adjustment for seasonal index   1.0260
3.8985 Two-tailed p-value  2(.0899)  .1798
b. The largest effect is in quarter 4; this seems reasonable p-value  .05, do not reject H0; conclude no indication
since retail sales are generally higher during October, that a difference exists
November, and December
4. a. H0: Median
15
52. a. Yes, a linear trend pattern appears to be present Ha: Median  15
b. A portion of the Minitiab regression output follows: b. n  9; number of plus signs  1
p-value  .0196
The regression equation is
Reject H0; bond mutual funds have lower median
Number Sold = 22.9 + 15.5 Year
6. n  48; z  1.88
c. Forecast in year 8 is or approximately 147 units p-value  .0301
Reject H0; conclude median  $56.2 thousand
54. b. The centered moving average values smooth out the
time series by removing seasonal effects and some of 8. a. n  15
the random variability; the centered moving average p-value  .0768
time series shows the trend in the data Do not reject H0; no significant difference for the pace
c. b. 25%, 68.8%; recommend larger sample
10. n  600; z = 2.41
Quarter Adjusted Seasonal Index p-value  .0160
1 0.899 Reject H0; significant difference, American Idol preferred
2 1.362
3 1.118 12. H0: Median for Additive 1  Median for Additive 2  0
4 0.621 Ha: Median for Additive 1  Median for Additive 2 0