Heizerom10ism04 141023093306 Conversion Gate02

CHAPTER 4F O R E C A S T I N G
27
C H A P T E R
Forecasting
DISCUSSION QUESTIONS
1.Qualitative models incorporate subjective factors into the
forecastingmodel.Qualitativemodelsareusefulwhensubjective
factors are important. When quantitative data are difficult to
obtain,qualitativemodelsmaybeappropriate.
2.Approaches are qualitative and quantitative. Qualitative is
relativelysubjective;quantitativeusesnumericmodels.
3.Shortrange (under3 months), mediumrange(3 monthsto
3years),andlongrange(over3years).
4.The steps that should be used to develop a forecasting
systemare:
(a)Determinethepurposeanduseoftheforecast
(b)Selecttheitemorquantitiesthataretobeforecasted
(c)Determinethetimehorizonoftheforecast
(d)Selectthetypeofforecastingmodeltobeused
(e)Gatherthenecessarydata
(f) Validatetheforecastingmodel
(g)Maketheforecast
(h)Implementandevaluatetheresults
5.Any three of: sales planning, production planning and
budgeting,cashbudgeting,analyzingvariousoperatingplans.
6.Thereisnomechanismforgrowthinthesemodels;theyare
built exclusively from historical demand values. Such methods
willalwayslagtrends.
7.Exponentialsmoothing isaweightedmovingaveragewhere
allpreviousvaluesareweightedwithasetofweightsthatdecline
exponentially.
8.MAD,MSE,andMAPEarecommonmeasuresofforecast
accuracy.Tofindthemoreaccurateforecastingmodel,forecast
witheachtoolforseveralperiodswherethedemandoutcomeis
known, and calculate MSE, MAPE, or MAD for each. The
smallererrorindicatesthebetterforecast.
9.TheDelphitechniqueinvolves:
(a)Assembling a group of experts in such a manner as to
preclude direct communication between identifiable
membersofthegroup
(b)Assemblingtheresponsesofeachexperttothequestions
orproblemsofinterest
(c)Summarizingtheseresponses
(d)Providingeachexpertwiththesummaryofallresponses
(e)Askingeachexperttostudythesummaryoftheresponses
andrespondagaintothequestionsorproblemsofinterest.
(f) Repeatingsteps(b)through(e)severaltimesasnecessary
toobtainconvergenceinresponses.Ifconvergencehas
not been obtained by the end of the fourth cycle, the
responsesatthattimeshouldprobablybeacceptedand
the process terminatedlittle additional convergence is
likelyiftheprocessiscontinued.
10.Atimeseriesmodelpredictsonthebasisoftheassumption
thatthefutureisafunctionofthepast,whereasanassociative
modelincorporates intothe modelthe variablesof factorsthat
mightinfluencethequantitybeingforecast.
11.Atimeseriesisasequenceofevenlyspaceddatapointswiththe
fourcomponentsoftrend,seasonality,cyclical,andrandomvariation.
12.When the smoothing constant, , is large (close to 1.0),
moreweightisgiventorecentdata;whenislow(closeto0.0),
moreweightisgiventopastdata.
13.Seasonalpatternsareoffixeddurationandrepeatregularly.
Cycles vary in length and regularity. Seasonal indices allow
genericforecaststobemadespecifictothemonth,week,etc.,
oftheapplication.
14.Exponentialsmoothingweighsallpreviousvalueswithaset
ofweightsthatdeclineexponentially.Itcanplaceafullweighton
themostrecentperiod(withanalphaof1.0).This,ineffect,isthe
nave approach, which places all its emphasis on last periods
actualdemand.
15.Adaptive forecasting refers to computer monitoring of
trackingsignalsandselfadjustmentifasignalpassesitspresent
limit.
16.Tracking signals alert the user of a forecasting tool to
periodsinwhichtheforecastwasinsignificanterror.
17.Thecorrelationcoefficientmeasuresthedegreetowhichthe
independentanddependentvariablesmovetogether.Anegative
value would mean that as X increases, Y tends to fall. The
variablesmovetogether,butmoveinoppositedirections.
18.Independentvariable(x)issaidtoexplainvariationsinthe
dependentvariable(y).
19.Nearlyeveryindustryhasseasonality.Theseasonalitymust
be filtered out for good mediumrange planning (of production
andinventory)andperformanceevaluation.
20.Therearemanyexamples.Demandforrawmaterialsand
componentpartssuchassteelortiresisafunctionofdemandfor
goodssuchasautomobiles.
Copyright2011PearsonEducation,Inc.publishingasPrenticeHall.
28
21.Obviously,aswegofartherintothefuture,itbecomesmore
difficulttomakeforecasts,andwemustdiminishourrelianceon
theforecasts.
*ActiveModels4.1,4.2,4.3,and4.4appearonourWebsite,
www.pearsonhighered.com/heizer.
ETHICAL DILEMMA
Thisexercise,derivedfromanactualsituation,dealsasmuchwith
ethicsaswithforecasting.Hereareafewpointstoconsider:
Noonelikesasystemtheydontunderstand,andmost
collegepresidentswouldfeeluncomfortablewiththisone.
Itdoesoffertheadvantageofdepoliticizingthefundsal
locationifusedwiselyandfairly.Buttodosomeansall
partiesmusthaveinputtotheprocess(suchassmoothing
constants)andalldataneedtobeopentoeveryone.
Thesmoothingconstantscouldbeselectedbyanagreed
uponcriteria(suchaslowestMAD)orcouldbebasedon
inputfromexpertsontheboardaswellasthecollege.
Abuseofthesystemistiedtoassigningalphasbasedon
whatresultstheyyield,ratherthanwhatalphasmakethe
mostsense.
Regressionisopentoabuseaswell.Modelscanusemany
yearsofdatayieldingoneresultor few yearsyieldinga
totallydifferentforecast.Selectionofassociativevariables
canhaveamajorimpactonresultsaswell.
Active Model Exercises*

ACTIVEMODEL4.1:MovingAverages
1.Whatdoesthegraphlooklikewhenn=1?
Theforecastgraphmirrorsthedatagraphbutoneperiod
later.
2.Whathappenstothegraphasthenumberofperiodsinthe
movingaverageincreases?
Theforecastgraphbecomesshorterandsmoother.
3.WhatvaluefornminimizestheMADforthisdata?
n=1(anaveforecast)
ACTIVEMODEL4.2:ExponentialSmoothing
1.Whathappenstothegraphwhenalphaequalszero?
Thegraphisastraightline.Theforecastisthesamein
eachperiod.
2.Whathappenstothegraphwhenalphaequalsone?
The forecast follows the same pattern as the demand
(exceptforthefirstforecast)butisoffsetbyoneperiod.This
isanaveforecast.
3.Generalizewhathappenstoaforecastasalphaincreases.
As alpha increases the forecast is more sensitive to
changesindemand.
4.2(a)No,thedataappeartohavenoconsistentpattern
(seepartdforgraph).
(b)
(c)
Year
Demand
3-year moving
3-year weighted
9.0
7.0
6.4
13.0
7.7
7.8
8.0
9.0
11.0
12.0
10.0
9.6
10
11 Forecas
t
13.0 9.0 11.0 7.0

11.0 11.0 11.3 11.0
10.9 12.2 10.5 10.6
9.0
8.4
4.Atwhatlevelofalphaisthemeanabsolutedeviation(MAD)
minimized?
alpha=.16
29
(d)The threeyear moving average appears to give better

results.
ACTIVEMODEL4.3:ExponentialSmoothingwith
TrendAdjustment
1.Scroll through different values for alpha and beta. Which

smoothingconstantappearstohavethegreatereffectonthegraph?
alpha
2.With betaset tozero, find thebest alphaand observethe
MAD. Now find the best beta. Observe the MAD. Does the
additionofatrendimprovetheforecast?
4.3
Year
Demand
Nave
Exp. Smoothing
9.0
7.0
6.4
5.0
9.0
7.4
10
11
Forecast
9.0 13.0 8.0 12.0 13.0 9.0 11.0 7.0

5.0 9.0 13.0 8.0 12.0 13.0 9.0 11.0
6.5 7.5 9.7 9.0 10.2 11.3 10.4 10.6
7.0
9.2
alpha=.11,MAD=2.59;betaabove.6changestheMAD
(byalittle)to2.54.
ACTIVEMODEL4.4:TrendProjections
1.Whatistheannualtrendinthedata?
10.54
2.Usethescrollbarsfortheslopeandintercepttodeterminethe
valuesthatminimizetheMAD.Arethesethesamevaluesthat
regressionyields?
No, they are not the same values. For example, an
interceptof57.81withaslopeof9.44yieldsaMADof7.17.
END-OF-CHAPTER PROBLEMS
4.1(a)
TEACHING NOTE: Notice how well exponential

smoothingforecaststhenave.
374+368+381
374.33pints
3
4.4 (a)FJuly FJune 0.2(Forecastingerror)
(b)
Week of
August 31
September
September
September
September
October 5
Navetrackstheupsanddownsbestbutlagsthedatabyone
period.Exponentialsmoothingisprobablybetterbecauseit
smoothesthedataanddoesnothaveasmuchvariation.
7
14
21
28
Pints
Used
360
389
410
381
368
374
Weighted
Moving Average
(b)FAugust FJuly 0.2(Forecastingerror)

41.6 0.2(45 41.6) 42.3
381 .1 = 38.1
368 .3 = 110.4
374 .6 = 224.4
372.9
Forecast
372.9
(c)
Theforecastis374.26.
42 0.2(40 42) 41.6
(c)Thebankingindustryhasagreatdealofseasonalityin
itsprocessingrequirements
Week of
Pints Forecast
August 31
September 7
September 14
September 21
360
389
410
381
360
360
365.8
374.64
Forecasti
ng
Error
0
29
44.2
6.36
September 28 368
375.912 7.912
October 5
374.3296.3296
4.5(a)
374
3,700 + 3,800
=3,750miles
2
(b)
Two-Year
Error
.20
0
5.8
8.84
1.272
Foreca
st
360
365.8
374.64
375.91
2
1.5824 374.32
96
.06592 374.26
36
30
Year
1
2
3
4
5
Mileage
Moving Average
3,000
4,000
3,400
3,800
3,700
300
100.
3
|Error|
4.5(c)Weighted2yearM.A.with.6weightformostrecentyear.
Year
3,500
3,700
3,600
Totals
MAD
Error
100
100
100
10
0
100
100
100
300
Mileage
Forecast
Error
|Error|
3,000
4,000
3,400
3,800
3,700
3,600
3,640
3,640
200
160
60
200
160
60
420
1
2
3
4
5
Forecastforyear6is3,740miles.
420
MAD 140
3
4.5(d)
Year
Mileage Forecast
1
2
3
4
5
3,000
4,000
3,400
3,800
3,700
Forecast
Error
3,000
3,000
3,500
3,450
3,625
Total
0
1,000
100
350
75
1,32
5
Error
New
= .50 Forecast
0
500
50
175
38
3,000
3,500
3,450
3,625
3,663
Theforecastis3,663miles.
4.6
January
February
March
April
May
June
July
August
September
October
November
December
Sum
Average
Y Sales
X Period
20
21
15
14
13
16
17
18
20
20
21
23
1
2
3
4
5
6
7
8
9
10
11
12
218
18.2
78
6.5
(a)
X2
1
4
9
16
25
36
49
64
81
100
121
144
XY
20
42
45
56
65
96
119
144
180
200
231
276
650
1,474
(b)[i]Nave
ThecomingJanuary=December=23
[ii]3monthmoving(20+21+23)/3=21.33
[iii]6monthweighted[(0.117)+(.118)
+(0.120)+(0.220)
+(0.221)+(0.323)]/1.0
=20.6
[iv]Exponentialsmoothingwithalpha=0.3
FOct 18 0.3(20 18) 18.6
FNov 18.6 0.3(20 18.6) 19.02
FDec 19.02 0.3(21 19.02) 19.6
FJan 19.6 0.3(23 19.6) 20.62 21
[v]Trend x 78,x 6.5, y = 218,y 18.17
xy nx y
x 2 nx 2
1474 (12)(6.5)(18.2) 54.4
b
0.38
650 12(6.5)2
143
a y bx
a 18.2 0.38(6.5) 15.73
b
Forecast=15.73+.38(13)=20.67,wherenextJanuary
isthe13thmonth.
(c)Onlytrendprovidesanequationthatcanextendbeyond
onemonth
4.7Present=Period(week)6.
1
1
1 1
a)So: F7 3 A6 4 A5 4 A4 6 A3

1.0
1
1
1
1
(52) + (63) + (48) + (70) = 56.76patients,
6
3
4
4
or57patients
1 1 1 1
where 1.0 = weights , , ,
3 4 4 6
b) Iftheweightsare20,15,15,and10,therewillbenochangein
theforecastbecausethesearethesamerelativeweightsasin
part(a),i.e.,20/60,15/60,15/60,and10/60.
c) If the weights are 0.4, 0.3, 0.2, and 0.1, then the forecast
becomes56.3,or56patients.
31
32
TableforProblem4.9(a,b,c)
Forecast
Two-Month
Month
January
February
March
April
May
June
July
August
September
October
November
December
Price per
Chip
$1.80
1.67
1.70
1.85
1.90
1.87
1.80
1.83
1.70
1.65
1.70
1.75
Moving
Average
1.735
1.685
1.775
1.875
1.885
1.835
1.815
1.765
1.675
1.675
(96 88 90)
91.3
3
(88 90)
(b)
89
2
(c)
ThreeMonth
Moving
Average
1.723
1.740
1.817
1.873
1.857
1.833
1.777
1.727
1.683
Totals
|Error|
Two-Month
ThreeMonth
Moving
Moving
Average
Average
.035
.165
.125
.005
.085
.005
.115
.115
.025
.075
.
.127
.160
.053
.073
.027
.133
.127
.027
.067
.793
4.8(a)
Temperatur 2 day
e
M.A.
|
(Error
Error|
)2
93
94
93
0.5 0.25
93.5
Absolute % Error
100(.5/93)
=
0.54%
95
93.5
96
94.0
88
95.5
90
92.0
1.5
2.25
=
100(1.5/95) 1.58%
2.0 4.00 100(2/96)
=
2.08%
7.5
=
56.25 100(7.5/88) 8.52%
2.0 4.00 100(2/90)
=
2.22%
13.
66.75
14.94
5
%
(c)MAD(twomonthmovingaverage)=.750/10=.075
MAD(threemonthmovingaverage)=.793/9=.088
Therefore, the twomonth moving average seems to have
performedbetter.
MAD=13.5/5=2.7
(d)MSE=66.75/5=13.35
(e)MAPE=14.94%/5=2.99%
4.9(a,b)Thecomputationsforboththetwoandthreemonth
averagesappearinthetable;theresultsappearinthe
figurebelow.
4.9
(d)TableforProblem4.9(d):
= .1
Month
January
February
March
April
May
June
July
August
Septembe
r
October
November
December
= .3
= .5
Price per
Chip
Forecast
|Error|
Forecast
|Error|
Forecast
|Error|
$1.80
1.67
1.70
1.85
1.90
1.87
1.80
1.83
1.70
$1.80
1.80
1.79
1.78
1.79
1.80
1.80
1.80
1.81
$.00
.13
.09
.07
.11
.07
.00
.03
.11
$1.80
1.80
1.76
1.74
1.77
1.81
1.83
1.82
1.82
$.00
.13
.06
.11
.13
.06
.03
.01
.12
$1.80
1.80
1.74
1.72
1.78
1.84
1.86
1.83
1.83
$.00
.13
.04
.13
.12
.03
.06
.00
.13
1.65
1.70
1.75
Totals
1.80
1.78
1.77
.15
.08
.02
$.8
6
$.072
1.79
1.75
1.73
.14
.05
.02
$.8
6
$.072
1.76
1.71
1.70
.11
.01
.05
$.81
MAD (total/12)
$.067
5
=.5ispreferable,usingMAD,to=.1or=.3.Onecould
4.10
Year
Demand
(a) 3-year moving
(b) 3-year
weighted
5.0 10.0 8.0

4.7 5.0 6.3
4.5 5.0 7.3
10
11
Forecast
7.0
7.7
7.8
9.0
8.3
8.0
12.0
8.0
8.3
14.0
9.3
10.0
15.0
11.7
12.3
13.7
14.0
(c)Theforecastsareaboutthesame.
4.11 (a) Year
Demand
Exp. Smoothing
4
5
6.0
4.7
4.0
5.1
5.0 10.0 8.0

4.8 4.8 6.4
10
11
Forecast
7.0
6.9
9.0
6.9
12.0
7.5
14.0
8.9
15.0
10.4
11.8
33
34
(b)|Error|=|ActualForecast|
Year
Exp. smoothing
1
1
2
1.3
3
1.1
4
0.2
5
5.2
6
1.6
7
0.1
8
2.1
9
4.5
10
5.1
11
4.6
MAD
2.4
ThesecalculationswerecompletedinExcel.CalculationsareslightlydifferentinExcelOMandPOMforWindowsdueto
roundingdifferences.
=12.3
MAD=6.2
4.12
Actual
Demand
Forecast
Demand
Monday
88
88
Tuesday
72
88
Wednesday
68
84
Thursday
48
80
Friday
Day
(c)Trendprojection:
Year
72
Answer
Ft=Ft1+(At1Ft1)
Demand
1
2
3
4
5
6
45
50
52
56
58
?
Trend Projection
42.6
42.6
42.6
42.6
42.6
42.6
+
+
+
+
+
+
3.2
3.2
3.2
3.2
3.2
3.2
1
2
3
4
5
6
=
=
=
=
=
=
45.8
49.0
52.2
55.4
58.6
61.8
Let=.25.LetMondayforecastdemand=88
Y a bX
F4=84+.25(6884)=844=80
F5=80+.25(4880)=808=72
Exponential
Absolute
Demand
Smoothing = 0.6
Deviation
45
41
4.0
50
41.0 + 0.6(4541) = 43.4
6.6
52
43.4 + 0.6(5043.4) = 47.4
4.6
56
47.4 + 0.6(5247.4) = 50.2
5.8
58
50.2 + 0.6(5650.2) = 53.7
4.3
?
53.7 + 0.6(5853.7) = 56.3
=25.3
MAD=5.06
Exponentialsmoothing,=0.9:
Year
1
2
3
4
5
6
Demand
Exponential
Smoothing = 0.9
Absolute
Deviation
45
50
52
56
58
?
41
41.0 + 0.9(4541) = 44.6
44.6 + 0.9(5044.6 ) = 49.5
49.5 + 0.9(5249.5) = 51.8
51.8 + 0.9(5651.8) = 55.6
55.6 + 0.9(5855.6) = 57.8
4.0
5.4
2.5
4.2
2.4
=18.5
MAD=3.7
(b)3yearmovingaverage:
Year
Demand
1
2
3
4
5
45
50
52
56
58
Three-Year
Moving Average
XY nXY
X 2 nX 2
a Y bX
4.13(a)Exponentialsmoothing,=0.6:
Year
1
2
3
4
5
6
0.8
1.0
0.2
0.6
0.6
=3.2
MAD=0.64
F2=88+.25(8888)=88+0=88
F3=88+.25(7288)=884=84
Absolute
Deviation
Absolute
Deviation
XY
X2
1
2
3
4
5
45
50
52
56
58
45
100
156
224
290
1
4
9
16
25
Then:X=15,Y=261,XY=815,X =55, X =3, Y =52.2

Therefore:
815 5 3 52.2
b
3.2
55 5 3 3
a 52.20 3.20 3 42.6
Y6 42.6 3.2 6 61.8
(d)Comparingtheresultsoftheforecastingmethodologies
forparts(a),(b),and(c).
Forecast Methodology
MAD
Exponential smoothing, = 0.6

Exponential smoothing, = 0.9
3-year moving average
Trend projection
5.06
3.7
6.2
0.64
Basedonameanabsolutedeviationcriterion,thetrendprojection
istobepreferredovertheexponentialsmoothingwith =0.6,
exponentialsmoothingwith=0.9,orthe3yearmovingaverage
forecastmethodologies.
4.14
(45 + 50 + 52)/3 = 49
(50 + 52 + 56)/3 =
52.7
(52 + 56 + 58)/3 =
55.3
7
5.3
Method1:
MAD:(0.20+0.05+0.05+0.20)/4=.125better
MSE:(0.04+0.0025+0.0025+0.04)/4=.021
Method2:
MAD:(0.1+0.20+0.10+0.11)/4=.1275
MSE:(0.01+0.04+0.01+0.0121)/4=.018better
4.15
Year
Sales
2005
2006
2007
2008
450
495
518
563
2009
584
2010
Forecast Three-Year
Moving Average
Absolute
Deviatio
n
(450 + 495 + 518)/3 =

487.7
(495 + 518 + 563)/3 =
525.3
(518 + 563 + 584)/3 =
555.0
75.3
58.7
=134
MAD=67
4.16
Year
Time Period X
Sales Y
X2
2005
2006
2007
2008
2009
1
2
3
4
5
450
495
518
563
584
1
4
9
16
25
450
990
1554
2252
2920
= 55
=
8166
XY
2610
X 3, Y 522
Y a bX
b
XY nXY
2
X nX
8166 (5)(3)(522) 336
33.6
55 (5)(9)
10
a Y bX 522 (33.6)(3) 421.2

y 421.2 33.6 x
y 421.2 33.6 6 622.8
Year
Sales
Forecast Trend
Absolute
Deviation
2005
2006
2007
2008
2009
2010
450
495
518
563
584
454.8
488.4
522.0
555.6
589.2
622.8
4.8
6.6
4.0
7.4
5.2
=28
MAD=5.6
4.17
Year
Sales
2005
2006
2007
2008
450
495
518
563
2009
2010
584
Forecast Exponential
Smoothing = 0.6
410.0
410 +
434 +
470.6
499.0
499 +
537.4
565.6
0.6(450 410) = 434.0

0.6(495 434) = 470.6
+ 0.6(518 470.6) =
0.6(563 499) = 537.4
+ 0.6(584 537.4) =
Absolute
Deviation
40.0
61.0
47.4
64.0
46.6
=259
MAD=51.8
35
36
Forecast Exponential
Year
Sales
2005
2006
2007
2008
450
495
518
563
2009
584
2010
Absolut
e
Deviatio
n
Smoothing = 0.9
410.0
410 +
446 +
490.1
515.2
515.2
558.2
558.2
581.4
0.9(450 410) = 446.0

0.9(495 446) = 490.1
+ 0.9(518 490.1) =
40.0
49.0
27.9
47.8
+ 0.9(563 515.2) =
25.8
Note:TousePOMforWindowstosolvethisproblem,aperiod0,
whichcontainstheinitialforecastandinitialtrend,mustbeadded.
+ 0.9(584 558.2) =
=190.5
MAD=38.1
(RefertoSolvedProblem4.1)
For =0.3,absolutedeviationsfor20052009are40.0,73.0,
74.1,96.9,88.8,respectively.SotheMAD=372.8/5=74.6.
MAD 0.3 74.6
MAD 0.6 51.8
MAD 0.9 38.1
Because it gives the lowest MAD, the smoothing constant of
=0.9givesthemostaccurateforecast.
4.18Weneedtofindthesmoothingconstant .Weknowin
generalthat Ft =Ft1 + (At1 Ft1); t= 2,3,4. Choose
either
t=3ort=4(t=2wontletusfindbecauseF2=50=50+
(5050)holdsforany).Letspickt=3.ThenF3=48=50+
(4250)
or48=50+4250
or2=8
So,.25=
NowwecanfindF5:F5=50+(4650)
F5=50+4650=504
For=.25,F5=504(.25)=49
Theforecastfortimeperiod5=49units.
4.19Trendadjustedexponentialsmoothing:=0.1,=0.2
Unadjust
Adjusted
ed
Month Income Forecast Trend Forecast |
Error 2
Error|
Februar
y
March
April
May
June
July
August
70.0
65.0
0.0
68.5
64.8
71.7
71.3
72.8
65.5
65.9
65.92
66.62
67.31
68.16
0.1
0.16
0.13
0.25
0.33
65
65.6
66.05
66.06
66.87
67.64
68.60
5.0 25.0
2.9
1.2
5.6
4.4
5.2
24.3
8.4
1.6
31.9
19.7
26.6
113.
2
MAD = 24.3/6 = 4.05, MSE = 113.2/6 = 18.87. Note that all

numbersarerounded.
37
4.20Trendadjustedexponentialsmoothing:=0.1,=0.8
4.23Students must determine the nave forecast for the four
Unadjusted
4.21 F5 A4 1 F4 T4 0.2
19 0.8 20.14
Month
Demand (y)
Forecast
3.8 16.11 19.91
Trend
February
70.0
65.0
0
March T5 F5 F4 168.5
65.5 17.82 0.4
T4 0.4 19.91
April
64.8
66.16
0.61
0.666.57
2.32 0.40.45
2.09
May
71.7
June
71.3
1.3967.49
0.84 1.390.82
2.23
July
72.8
68.61
1.06
2.23 22.14
Totals FIT5 F5 T5 19.91
419.1
Average
F A5 1 69.85
F5 T5 0.2 24 0.8 22.14
August6Forecast
Adjusted
months.ThenaveforecastforMarchistheFebruaryactualof83,
Forecast
Error |Error| Error2
etc.
65.0
65.9
66.77
67.02
68.31
69.68
5.00
2.60
1.97
4.68
2.99
3.12
5.0
2.6
1.97
4.68
2.99
3.12
20.3
6
3.39
(MAD)
16.42
71.30
2.74
(Bias)
25.00
6.76
3.87
21.89
8.91
9.76
76.19
12.70
(MSE)
4.8 17.71 22.51

Based upon the MSE criterion, the exponential smoothing with = 0.1, = 0.8 is to be preferred
overtheexponentialsmoothingwith
=0.1,
=0.2.
ItsMADof3.39is
T6 F6 F5 1 T5 0.4
22.51
19.91
0.6ItsMSEof12.70islower.
2.23
alsolowerthanthatinProblem4.19.
0.4 2.6 1.34
(a)
Actual Forecas |Error| |% Error|
1.04 1.34 2.38
FIT6 F6 T6 22.51 2.38 24.89

4.22 F7 A6 (1 )( F6 T6 ) (0.2)(21) (0.8)(24.89)
4.2 19.91 24.11
T7 ( F7 F6 ) (1 )T6 (0.4)(24.11 22.51)
(0.6)(2.38) 2.07
FIT7 F7 T7 24.11 2.07 26.18
March
101
120
19
April
96
114
18
May
89
110
21
June
108
108
0
58
100
(19/101)
=
18.81%
100
(18/96)
=
18.75%
100
(21/89)
=
23.60%
100 (0/108) =
0%
61.16%
58
14.5
4
61.16%
MAPE(formanagement)
15.29%
4
F8 A7 (1 )( F7 T7 ) (0.2)(31)
MAD(formanagement)
(0.8)(26.18) 27.14
T8 F8 F7 1 T7 0.4 27.14 24.11
0.6 2.07 2.45
FIT8 F8 T8 27.14 2.45 29.59

F9 A8 1 F8 T8 0.2 28
0.8 29.59 29.28
T9 F9 F8 1 T8 0.4 29.28 27.14
0.6 2.45 2.32
FIT9 F9 T9 29.28 2.32 31.60
Actual Nave |Error| |% Error|
(b)
March
101
83
18
April
May
96
89
101
96
5
7
June
108
89
19
4
9
100
(18/101)
=
17.82%
100 (5/96) = 5.21%
100
(7/89)
=
7.87%
100
(19/108)
=
17.59%
48.49
%
38
(d) R=.82isthecorrelationcoefficient,andR 2 =.68

means 68% of the variation in sales can be explained by TV
appearances.
49
12.25
4
48.49%
MAPE(fornave)
12.12%.
4
Naveoutperformsmanagement.
MAD(fornave)
4.25
(c)MADforthemanagerstechniqueis14.5,whileMADforthe
naveforecastisonly12.25.MAPEsare15.29%and12.12%,
respectively.Sothenavemethodisbetter.
4.24(a)Graphofdemand
Theobservationsobviouslydonotformastraightlinebutdotend
toclusteraboutastraightlineovertherangeshown.
(b)Leastsquaresregression:
Month
January
February
March
April
Totals
Averages
Number of
Accidents
(y)
30
40
60
90
220
y = 55
x
1
2
3
4
10
x = 2.5
Y a bX
b
XY nXY
X 2 nX 2
a Y bX
Assume
Appearances
X
Demand Y
X2
Y2
XY
3
4
7
6
8
5
9
3
6
7
5
10
7
?
9
16
49
36
64
25
9
36
49
25
100
49
9
24
49
30
80
35
X =33, Y =38, XY =227, X2 =199, X =5.5, Y =6.33.

Therefore:
227 6 5.5 6.333
1.0286
199 6 5.5 5.5
a 6.333 1.0286 5.5 .6762
Y .676 1.03 x (rounded)
b
Thefollowingfigureshowsboththedataandtheresultingequation:
(c) If there are nine performances by Stone Temple

Pilots,theestimatedsalesare:
Y9 .676 1.03 9 .676 9.27 9.93 drums
10drums
20,000
1, 250
16
6,000
Averageoverspring:
1,500
4
1,500
Springindex:
1.2
1, 250
5,600
Answer:
(1.2) 1,680 sailboats
4
Averageoverallseasons:
xy
x2
30
80
180
360
650
1
4
9
16
30
xy n x y
650 4(2.5)(55)
x nx
100
20
5
a y bx
55 (20)(2.5)
5
30 4(2.5)
650 550
30 25
Theregressionlineisy=5+20x.TheforecastforMay(x=5)is
y=5+20(5)=105.
4.26
Averag
e
Season
Year1 Year2
Average
Season Year3
Seaso Deman DemanYear1Year2 Deman
al
Deman
n
d
d
Demand
d
Index
d
Fall
Winter
Spring
Summe
r
200
350
150
300
250
300
165
285
225.0
325.0
157.5
292.5
AverageYr1 to Yr2

Demandforseason
250
250
250
250
0.90
1.30
0.63
1.17
270
390
189
351
Yr1 Demand Yr2 Demand

2
SumofAveYr1 to Yr2 Demand

4
AverageYr1 to Yr2 Demand
Seasonalindex=
AverageSeasonalDemand
NewAnnualDemand
Yr3
Seasonalindex
4
1200
Seasonalindex
4
Averageseasonaldemand
4.27
Winter
2006
2007
2008
2009
1,400
1,200
1,000
900
4,500
Spring
Summer
1,500
1,400
1,600
1,500
6,000
1,000
2,100
2,000
1,900
7,000
Fall
600
750
650
500
2,500
4.28
Quarte 2007 2008

r
Winter
Spring
Summer
Fall
73
104
168
74
65
82
124
52
2009
89
146
205
98
Average
Averag Quarterl Season
e
y
al
Deman Demand Index
d
75.67
110.67
165.67
74.67
106.67
106.67
106.67
106.67
0.709
1.037
1.553
0.700
39
4.32(a)
40
4.292011is25yearsbeyond1986.Therefore,the2011quarter
numbersare101through104.
(1)
Quarter
(2)
Quarter
Number
(3)
Forecast
(77 + .
43Q)
Winter
Spring
Summer
Fall
101
102
103
104
120.43
120.86
121.29
121.72
96.344
132.946
169.806
85.204
(c) Y=36+4.3(90)=423
Letx1, x2 , K , x6 bethepricesand y1, y2 , K , y6
bethenumbersold.
Y
6
Averageprice=3.2583
(3)
2
xi =67.1925
(4)
xi yi nx y
i 1
nx
19,560
920
2,880 5(3)(180)
2
2,880 2,700
55 45
55 5(3)
180
18
10
a 180 3(18) 180 54 126
y 126 18 x
Theny=a+bx,wherey=numbersold,x=price,and
x 2i
1,280
4.33(a)Seethetablebelow.
(2)
xi yi 9,783
i1
i 1
6
60
(a)28
(b)43
(1)
Averagenumbersold=550
i 1
6
256
144
324
196
4.34Y=7.5+3.5X1+4.5X2+2.5X3
i 1
i 1
5,280
3,240
6,840
4,200
(b)MSE=160/5=32
(c)MAPE=13.23%/5=2.65%
xi
6
330
270
380
300
(b)Iftheforecastisfor20guests,thebarsalesforecast
is 50 + 18(20) = $410. Each guest accounts for an
additional$18inbarsales.
(b)Y=36+4.3(80)=380
yi
16
12
18
14
60
15
4
1,280
y
320
4
xy nx y 19,560 4(15)(320) 360
b
18
20
x 2 nx 2
920 4(15)2
a y bx 320 18(15) 50
Y 50 18 x
(a) Y=36+4.3(70)=337
xy
4.30GivenY=36+4.3X
4.31
Soatx=2.80,y=1,454.6277.6($2.80)=677.32.Nowroundto
thenearestinteger:Answer:677lattes.
(5)
(4)
Adjusted
Seasonal Forecast
Factor
[(3) (4)]
.8
1.1
1.4
.7
x2
(9, 783) 6(3.25833)(550)

67.1925 6(3.25833)2
969.489
277.6
3.49222
a y bx 550 [(277.6)(3.25)] 1, 454.6
TableforProblem4.33
Year
(x)
Totals
Transistors
(y)
xy
x2
126 + 18x
Error
Error 2
140
140
144
16
160
320
162
190
570
180
10
100
200
800
16
198
210
1,050
25
216
36
2,80
0
5
5
1
5
x = 3
90
0
y = 180
16
0
|% Error|
100 (4/140) =
2.86%
100 (2/160) =
1.25%
100 (10/190) =
5.26%
100 (2/200) =
1.00%
100 (6/210) =
2.86%
13.23
%
(c)58
4.35(a) Y =13,473+37.65(1860)=83,502
(b)Thepredictedsellingpriceis$83,502,butthisisthe
averagepriceforahouseofthissize.Thereareother
Cumulative error = 14.05; MAD = 5Tracking = 14.05/5

2.82
factors besides square footage that will impact the

selling price of a house. If such a house sold for
$95,000,thentheseotherfactorscouldbecontributing
totheadditionalvalue.
(c)Someotherquantitativevariableswouldbeageofthe
house,numberofbedrooms,sizeofthelot,andsizeof
thegarage,etc.
4.38(a)leastsquaresequation:Y=0.158+0.1308X
(b)Y= 0.158+0.1308(22)=2.719million
(c)coefficientofcorrelation=r=0.966
coefficientofdetermination=r2=0.934
4.39
(d)Coefficient of determination = (0.63)2 = 0.397. This

meansthatonlyabout39.7%ofthevariabilityinthe
salespriceofahouseisexplainedbythisregression
model that only includes square footage as the
explanatoryvariable.
4.36(a)Given:Y=90+48.5X1+0.4X2where:
Y expectedtravelcost
X1 numberofdaysontheroad
X2 distancetraveled,inmiles
r 0.68(coefficientofcorrelation)
Year X
1
2
3
4
5
6
7
8
9
10
55
Patients Y
X2
Y2
XY
36
33
40
41
40
55
60
54
58
61
478
1
4
9
16
25
36
49
64
81
100
385
1,296
1,089
1,600
1,681
1,600
3,025
3,600
2,916
3,364
3,721
23,89
2
36
66
120
164
200
330
420
432
522
610
2,900
Given:Y=a+bXwhere:
If:
XY nXY
Number of days on the road X1 = 5 and distance

traveledX2=300
then:
a Y bX
Y=90+48.55+0.4300=90+242.5+120=452.5
Therefore,theexpectedcostofthetripis$452.50.
(b)The reimbursement request is much higher than
predictedbythemodel.Thisrequestshouldprobably
bequestionedbytheaccountant.
(c)Anumberofothervariablesshouldbeincluded,suchas:
1.thetypeoftravel(airorcar)
2900 10 5.5 47.8
Inaddition,thecorrelationcoefficientof0.68isnotexceptionally
high.Itindicatesthatthemodelexplainsapproximately46%of
the overall variation in trip cost. This correlation coefficient
wouldsuggestthatthemodelisnotaparticularlygoodone.
4.37(a,b)
20
20
20.5
24.25
30.63
27.81
28.41
32.20
27.11
26.05
Error Running sum |error|

0.00
1.00
7.50
12.75
5.63
1.19
7.59
10.20
2.10
1.95
2900 2629 271
3.28
385 302.5 82.5
andY=29.76+3.28X.For:
4.other transportation costscab, limousine, special

tolls,orparking
20
21
28
37
25
29
36
22
25
28
385 10 5.5
a 47.8 3.28 5.5 29.76
X 11:Y 29.76 3.28 11 65.8

X 12:Y 29.76 3.28 12 69.1
3.costsofentertainingcustomers
Perio Demand Forecas

d
t
X 2 nX 2
andX=55, Y=478, XY=2900, X 2 =385, Y 2 =23892,

X 5.5,Y 47.8, Then:
2.conferencefees,ifany
1
2
3
4
5
6
7
8
9
10
41
0.00
1.00
8.50
21.25
15.63
16.82
24.41
14.21
12.10
14.05
Therefore:
Year1165.8patients
Year1269.1patients
The model seems to fit the data pretty well. One should,
however,bemorepreciseinjudgingtheadequacyofthemodel.
Twopossibleapproachesarecomputationof(a)thecorrelation
coefficient, or(b)themeanabsolutedeviation. The correlation
coefficient:
0.00
1.00
7.50
12.75
5.63
1.19
7.59
10.20
2.10
1.95
MAD
5.00
42
n XY X Y
n X 2 X 2
2
n Y 2 Y
10 2900 55 478
10 385 552 10 23892 478
2
29000 26290
3850 3025 238920 228484
2710
825 10436
2710
0.924
2934.3
r 2 0.853
Thecoefficientofdeterminationof0.853isquiterespectable
indicatingouroriginaljudgmentofagoodfitwasappropriate.
Year Patients
X
Y
1
36
33
40
41
40
55
60
54
58
10
61
Trend
Forecast
Absolute
Deviati Deviation
on
29.8 + 3.28 1 =
33.1
29.8 + 3.28 2 =
36.3
29.8 + 3.28 3 =
39.6
29.8 + 3.28 4 =
42.9
29.8 + 3.28 5 =
46.2
29.8 + 3.28 6 =
49.4
29.8 + 3.28 7 =
52.7
29.8 + 3.28 8 =
56.1
29.8 + 3.28 9 =
59.3
29.8 + 3.28 10 =
62.6
2.9
2.9
3.3
3.3
0.4
0.4
X 2 nX 2
and X = 854, Y = 478, XY = 42558.6, X2 = 76129.9,

Y 2 =23892, X =85.4, Y =47.8.Then:
b
42558.6 10 85.4 47.8
76129.9 10 85.42
1737.4
0.543
3197.3
a 47.8 0.543 85.4 1.43
1.9
6.2
6.2
5.6
5.6
andY=1.43+0.543X
7.3
7.3
For:
2.1
2.1
1.3
1.3
1.6
1.6
The MAD is 3.26this is approximately 7% of the average

numberofpatientsand10%oftheminimumnumberofpatients.
Wealsoseeabsolutedeviations,foryears5,6,and7intherange
5.67.3. The comparison of the MAD with the average and
minimum number of patients and the comparatively large
deviationsduringthemiddleyearsindicatethattheforecastmodel
is not exceptionally accurate. It is more useful for predicting
generaltrendsthantheactualnumberofpatientstobeseenina
specificyear.
XY nXY
a Y bX
1.9
=32.6
MAD=3.26
43
42558.6 40821.2
76129.9 72931.6
X 131.2 : Y 1.43 0.543(131.2) 72.7

X 90.6 : Y 1.43 0.543(90.6) 50.6
Therefore:
Crimerate=131.272.7patients
Crimerate=90.650.6patients
NotethatroundingdifferencesoccurwhensolvingwithExcel.
4.41
(a)Itappearsfromthefollowinggraphthatthepointsdo
Alsonotethatacrimerateof131.2isoutsidetherangeofthe
scatteraroundastraightline.
datasetusedtodeterminetheregressionequations,socautionis
advised.
4.40
Year
Crime Patient
s
Rate X
Y
X2
Y2
58.3
36
3,398.9 1,296
61.1
33
3,733.2 1,089
73.4
40
5,387.6 1,600
75.7
41
5,730.5 1,681
81.1
40
6,577.2 1,600
89.0
55
7,921.0 3,025
101.1
60
10,221.2 3,600
94.8
54
8,987.0 2,916
103.3
58
10,670.9 3,364
10
116.2
61
13,502.4 3,721
Column Totals
854.0
478
76,129.
9
Given:Y=a+bXwhere
23,89
2
XY
2,098.8
2,016.3
2,936.0
3,103.7
3,244.0
4,895.0
6,066.0
5,119.2
5,991.4
7,088.2
42,558.
6
(b)Developingtheregressionrelationship,wehave:
(Summe Tourists Ridership
r
months) (Millions) (1,000,000s
)
Year
(X)
(Y)
1
2
3
4
5
6
7
8
9
10
11
12
7
2
6
4
14
15
16
12
14
20
15
7
1.5
1.0
1.3
1.5
2.5
2.7
2.4
2.0
2.7
4.4
3.4
1.7
Given:Y=a+bXwhere:
b
XY nXY
X 2 nX 2
a Y bX
X2
Y2
XY
49
4
36
16
196
225
256
144
196
400
225
49
2.25
1.00
1.69
2.25
6.25
7.29
5.76
4.00
7.29
19.36
11.56
2.89
10.5
2.0
7.8
6.0
35.0
40.5
38.4
24.0
37.8
88.0
51.0
11.9
44
and X = 132, Y = 27.1, XY = 352.9, X2 = 1796,

Y 2 =71.59, X =11, Y =2.26.
Then:
352.9 298.3 54.6
0.159
1796 1452
344
1796 12 112
a 2.26 0.159 11 0.511
b
and
352.9 12 11 2.26
Y=0.511+0.159X
(c)Given a tourist population of 10,000,000, the model
predictsaridershipof:
Y=0.511+0.15910=2.101,or2,101,000persons.
(d)If there are no tourists at all, the model predicts a
ridershipof0.511,or511,000persons.Onewouldnot
place much confidence in this forecast, however,
because the number of tourists (zero) is outside the
rangeofdatausedtodevelopthemodel.
(e)Thestandarderroroftheestimateisgivenby:
Syx
Y 2 a Y b XY
n2
71.59 0.511 27.1 0.159 352.9
12 2
71.59 13.85 56.11

.163
10
.404(roundedto.407inPOMforWindowssoftware)
(f) Thecorrelationcoefficientandthecoefficientofdeter
minationaregivenby:
r
n XY X Y
n X 2 X 2 n Y 2 Y 2
12 352.9 132 27.1
Apr.
May.
Jun.
35
42
50
Oct.
Nov.
Dec.
45
35
38
29
Both history and forecast for the next year are shown in the
accompanyingfigure:
12 1796 1322 12 71.59 27.1

2
4234.8 3577.2
21552 17424 859.08 734.41
657.6
657.6
0.917
4128 124.67 64.25 11.166
andr 2 0.840
4.42(a)This problem gives students a chance to tackle a
realisticprobleminbusiness,i.e.,notenoughdatato
make a good forecast. As can be seen in the
accompanyingfigure,thedatacontainsbothseasonal
andtrendfactors.
4.43(a)and(b)Seethefollowingtable:
Actual Smoothed
Averagingmethodsarenotappropriatewithtrend,seasonal,
orotherpatternsinthedata.Movingaveragessmoothoutseason
ality. Exponential smoothing can forecast January next year,
butnotfarther.Becauseseasonalityisstrong,anavemodelthat
studentscreateontheirownmightbebest.
(b)Onemodelmightbe:Ft+1=At11
Thatisforecastnextperiod=actualoneyearearlier toaccount
for
seasonality.Butthisignoresthetrend.
Oneverygoodapproachwouldbetocalculatetheincrease
fromeachmonthlastyeartoeachmonththisyear,sumall12
increases, and divide by 12. The forecast for next year would
equalthe valuefor thesame monththis yearplus theaverage
increaseoverthe12monthsoflastyear.
(c)Usingthismodel,theJanuaryforecastfornextyearbecomes:
148
F25 17
17 12 29
12
where148=totalmonthlyincreasesfromlastyeartothisyear.
Theforecastsforeachofthemonthsofnextyearthenbecome:
Jan.
Feb.
Mar.
29
26
32
July.
Aug.
Sep.
56
53
45
Week
Value
Value
A(t)
Ft ( =
0.2)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
50
35
25
40
45
35
20
30
35
20
15
40
55
35
25
55
55
40
35
60
75
50
40
65
+50.0
+50.0
+47.0
+42.6
+42.1
+42.7
+41.1
+36.9
+35.5
+35.4
+32.3
+28.9
+31.1
+35.9
+36.7
+33.6
+37.8
+41.3
+41.0
+39.8
+43.9
+50.1
+50.1
+48.1
+51.4
Smoothe
d
Forecast Value Forecas
t
Error
Ft ( =
Error
0.6)
+0.0
15.0
22.0
2.6
2.9
7.7
21.1
6.9
0.5
15.4
17.3
+11.1
+23.9
0.9
10.7
+21.4
+17.2
1.3
6.0
+20.2
+31.1
0.1
10.1
+16.9
MAD = 11.8
+50.0
+50.0
+41.0
+31.4
+36.6
+41.6
+37.6
+27.1
+28.8
+32.5
+25.0
+19.0
+31.6
+45.6
+39.3
+30.7
+45.3
+51.1
+44.4
+38.8
+51.5
+65.6
+56.2
+46.5
+57.6
+0.0
15.0
16.0
+8.6
+8.4
6.6
17.6
+2.9
+6.2
12.5
10.0
+21.0
+23.4
10.6
14.3
+24.3
+9.7
11.1
9.4
+21.2
+23.5
15.6
16.2
+18.5
MAD = 13.45
(c)Studentsshouldnotehowstablethesmoothedvalues
arefor=0.2.Whencomparedtoactualweek25calls
of85,thesmoothingconstant,=0.6,appearstodoa
slightlybetterjob.Onthebasisofthestandarderror
of the estimate and the MAD, the 0.2 constant is
better.However,othersmoothingconstantsneedtobe
examined.
46
4.44
Week
Actual Value
At
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Smoothed
Value
Ft ( = 0.3)
Trend Estimate
Tt ( = 0.2)
FITt
50.000
50.000
45.500
38.720
37.651
38.543
36.555
30.571
28.747
29.046
25.112
20.552
24.526
32.737
33.820
31.649
38.731
44.664
44.937
43.332
49.209
58.470
58.445
54.920
59.058
0.000
0.000
0.900
2.076
1.875
1.321
1.455
2.361
2.253
1.743
2.181
2.657
1.331
0.578
0.679
0.109
1.503
2.389
1.966
1.252
2.177
3.594
2.870
1.591
2.100
50.000
50.000
44.600
36.644
35.776
37.222
35.101
28.210
26.494
27.303
22.931
17.895
23.196
33.315
34.499
31.758
40.234
47.053
46.903
44.584
51.386
62.064
61.315
56.511
61.158
50.000
35.000
25.000
40.000
45.000
35.000
20.000
30.000
35.000
20.000
15.000
40.000
55.000
35.000
25.000
55.000
55.000
40.000
35.000
60.000
75.000
50.000
40.000
65.000
To evaluate the trend adjusted exponential smoothing model,

actualweek25callsarecomparedtotheforecastedvalue.The
model appears to be producing a forecast approximately mid
rangebetweenthatgivenbysimpleexponentialsmoothingusing
=0.2and =0.6.Trendadjustmentdoesnotappeartogive
anysignificantimprovement.
Forecast
Forecas
t
Error
0.000
15.000
19.600
3.356
9.224
2.222
15.101
1.790
8.506
7.303
7.931
22.105
31.804
1.685
9.499
23.242
14.766
7.053
11.903
15.416
23.614
12.064
21.315
8.489
4.46(a)
Tracking signal
4.45
( At Ft )
t 1
Month
At
Ft
May
June
July
August
Septemb
er
October
Novembe
r
Decembe
r
100
80
110
115
105
100
104
99
101
104
110
125
104
105
120
109
87
10.875
8
39
3.586
10.875
MAD
|At Ft |
0
24
11
14
(At Ft )
0
24
11
14
20
20
11
11
Sum: 87
Sum: 39
So:MAD:
Column totals
X2
Y2
421 2.90
177241 8.41
377 2.93
142129 8.58
585 3.00
342225 9.00
690 3.45
476100 11.90
608 3.66
369664 13.40
390 2.88
152100 8.29
415 2.15
172225 4.62
481 2.53
231361 6.40
729 3.22
531441 10.37
501 1.99
251001 3.96
613 2.75
375769 7.56
709 3.90
502681 15.21
366 1.60
133956 2.56
6885 36.96
385789 110.2
3
6
Given:Y=a+bXwhere:
XY
1220.9
1104.6
1755.0
2380.5
2225.3
1123.2
892.3
1216.9
2347.4
997.0
1685.8
2765.1
585.6
20299.
5
XY nXY
X 2 nX 2
a Y bX
47
48
and X =6885, Y =36.96, XY =20299.5, X 2 =3857893,

Y 2 =110.26, X =529.6, Y =2.843,Then:
b
20299.5 13 529.6 2.843
3857893 13 529.62
726
0.0034
211703
a 2.84 0.0034 529.6 1.03
20299.5 19573.5
3857893 3646190
Amits MAD for exponential smoothing (16.11) is lower

than that of Barbaras moving average (19.17). So his
forecastseemstobebetter.
andY=1.03+0.0034X
Asanindicationoftheusefulnessofthisrelationship,wecan
calculatethecorrelationcoefficient:
r
n XY X Y
n X 2 X 2 n Y 2 Y 2
13 20299.5 6885 36.96

13 3857893 68852 13 110.26 36.96
2
263893.5 254469.6
50152609 47403225 1433.4 1366.0
9423.9
2749384 67.0
9423.9
0.692
1658.13 8.21
r 2 0.479
A correlation coefficient of 0.692 is not particularly high. The
2
coefficientofdetermination,r ,indicatesthatthemodelexplains
only47.9%oftheoverallvariation.Therefore,whilethemodel
doesprovideanestimateofGPA,thereisconsiderablevariation
inGPA,whichisasyetunexplained.For
(b)X 350:Y 1.03 0.0034 350 2.22
(c)X 800:Y 1.03 0.0034 800 3.75
Note:Whensolvingthisproblem,caremustbetakentointerpret
significantdigits.AlsonotethatX =800isoutsidetherangeofthe
datasetusedtodeterminetheregressionrelationship,socautionis
advised.
4.47(a)Thereisnotastronglineartrendinsalesovertime.
(b,c)Amitwantstoforecastbyexponentialsmoothing(setting
FebruarysforecastequaltoJanuaryssales)withalpha
0.1.Barbarawantstousea3periodmovingaverage.
January
February
March
April
May
Sales
Amit
400
380
410
375
405
400
398
399.2
396.8
Barba Amit Error Barbara

ra
Error
396.67
388.33
MAD =
20.0
12.0
24.2
8.22
16.1
1
21.67
16.67
19.1
7
(d) Note that Amit has more forecast observations, while

Barbarasmovingaveragedoesnotstartuntilmonth4.Also
note that the MAD for Amit is an average of 4 numbers,
whileBarbarasisonly2.
4.48(a)
(Continued)
Quarter Contracts Sales Y

X
1
2
3
4
5
6
7
8
Totals
153
172
197
178
185
199
205
226
1,51
5
189.375
Average
8
10
15
9
12
13
12
16
95
23,409
29,584
38,809
31,684
34,225
39,601
42,025
51,076
290,41
3
64
100
225
81
144
169
144
256
1,183
XY
1,224
1,720
2,955
1,602
2,220
2,587
2,460
3,616
18,384
11.875
b = (18384 8 189.375 11.875)/(290,413 8 189.375

189.375)=0.1121
a=11.8750.1121189.375=9.3495
Sales(y)=9.349+0.1121(Contracts)
(b)
r (8 18384 1515 95)
((8 290,413 15152 )(8 1183 952 ))
0.8963
Sxy 1183 ( 9.3495 95) (0.112 18384 / 6) 1.3408
r 2 .8034
4.49(a)
Year
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
49
Method Exponential Smoothing

0.6 =
Deposits
Forecast
|Error|
Error2
(Y )
0.25
0.24
0.24
0.26
0.25
0.30
0.31
0.32
0.24
0.26
0.25
0.33
0.50
0.95
1.70
2.30
2.80
2.80
2.70
3.90
4.90
5.30
6.20
4.10
4.50
6.10
7.70
10.10
15.20
0.25
0.25
0.244
0.241
0.252
0.251
0.280
0.298
0.311
0.268
0.263
0.255
0.300
0.420
0.738
1.315
1.906
2.442
2.656
2.682
3.413
4.305
4.90
5.680
4.732
4.592
5.497
6.818
8.787
0.00
0.01
0.004
0.018
0.002
0.048
0.029
0.021
0.071
0.008
0.013
0.074
0.199
0.529
0.961
0.984
0.893
0.357
0.043
1.217
1.486
0.994
1.297
1.580
0.232
1.507
2.202
3.281
6.412
0.00
0.0001
0.0000
0.0003
0.00
0.0023
0.0008
0.0004
0.0051
0.0000
0.0002
0.0055
0.0399
0.2808
0.925
0.9698
0.7990
0.1278
0.0018
1.4816
2.2108
0.9895
1.6845
2.499
0.0540
2.2712
4.8524
10.7658
41.1195
50
4.49(a)(Continued)
Year
Method Exponential Smoothing

0.6 =
Deposits Forecast
|Error|
Error2
(Y )
30
31
32
33
34
35
36
18.10
24.10
25.60
30.30
36.00
31.10
31.70
12.6350
15.9140
20.8256
23.69
27.6561
32.6624
31.72
37
38
39
40
41
42
43
44
TOTALS
38.50
47.90
49.10
55.80
70.10
70.90
79.10
94.00
787.30
31.71
35.784
43.0536
46.6814
52.1526
62.9210
67.7084
74.5434
AVERAGE
17.8932
Next period forecast = 86.2173
5.46498
29.8660
8.19
67.01
4.774
22.7949
6.60976
43.69
8.34390
69.62
1.56244 2.44121
0.024975
0.000624
6.79
46.1042
12.116
146.798
6.046
36.56
9.11856
83.1481
17.9474
322.11
7.97897
63.66
11.3916
129.768
19.4566
378.561
150.3
1,513.2
2
3.416
34.39
(MAD)
(MSE)
Standard error =
6.07519
Method Linear Regression (Trend Analysis)

Year
Period (X ) Deposits (Y ) Forecast
Error2
1
2
3
4
5
6
7
8
9
10
11
12
1
2
3
4
5
6
7
8
9
10
11
12
0.25
0.24
0.24
0.26
0.25
0.30
0.31
0.32
0.24
0.26
0.25
0.33
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
0.50
0.95
1.70
2.30
2.80
2.80
2.70
3.90
4.90
5.30
6.20
4.10
4.50
6.10
7.70
10.10
15.20
17.330
15.692
14.054
12.415
10.777
9.1387
7.50
5.8621
4.2238
2.5855
0.947
0.691098
2.329
3.96769
5.60598
7.24427
8.88257
10.52
12.1592
13.7974
15.4357
17.0740
18.7123
20.35
21.99
23.6272
25.2655
26.9038
28.5421
309.061
253.823
204.31
160.662
121.594
89.0883
61.0019
38.2181
19.9254
8.09681
1.43328
0.130392
3.34667
9.10642
15.2567
24.4458
36.9976
59.6117
89.4756
97.9594
111.0
138.628
156.558
264.083
305.862
307.203
308.547
282.367
178.011
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
TOTALS
990.0
0
AVERAGE 22.50
18.10
24.10
25.60
30.30
36.00
31.10
31.70
38.50
47.90
49.10
55.80
70.10
70.90
79.10
94.00
787.30
17.893
30.18
31.8187
33.46
35.0953
36.7336
38.3718
40.01
41.6484
43.2867
44.9250
46.5633
48.2016
49.84
51.4782
53.1165
145.936
59.58
61.73
22.9945
0.5381
52.8798
69.0585
9.91266
21.2823
17.43
85.3163
479.54
443.528
762.964
1,671.46
7,5
59.95
171.817
(MSE)
Method Least squaresSimple Regression on

GSP
a
b
17.636 13.5936
Coefficients GSP Deposit
:
s
Year
(X)
(Y ) Forecast |Error|
Error2
1
2
3
4
5
6
7
8
9
0.40
0.40
0.50
0.70
0.90
1.00
1.40
1.70
1.30
0.25
0.24
0.24
0.26
0.25
0.30
0.31
0.32
0.24
10
11
12
13
14
15
16
17
1.20
1.10
0.90
1.20
1.20
1.20
1.60
1.50
0.26
0.25
0.33
0.50
0.95
1.70
2.30
2.80
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
1.60
1.70
1.90
1.90
2.30
2.50
2.80
2.90
3.40
3.80
4.10
4.00
4.00
3.90
3.80
3.80
2.80
2.70
3.90
4.90
5.30
6.20
4.10
4.50
6.10
7.70
10.10
15.20
18.10
24.10
25.60
30.30
12.198 12.4482
12.198 12.4382
10.839 11.0788
8.12
8.38
5.4014 5.65137
4.0420 4.342
1.395451.08545
5.473545.15354
0.036086 0.203914
1.3233 1.58328
2.6826 2.93264
5.4014 5.73137
1.3233 1.82328
1.3233 2.27328
1.3233 3.02328
4.114181.81418
2.75481
0.045186
4.114181.31418
5.473542.77354
8.192274.29227
8.192273.29227
13.6297 8.32972
16.3484 10.1484
20.4265 16.3265
21.79
17.29
28.5827 22.4827
34.02
26.32
38.0983 27.9983
36.74
21.54
36.74
18.64
35.3795 11.2795
34.02
8.42018
34.02
3.72018
154.957
154.71
122.740
70.226
31.94
18.8530
1.17820
26.56
0.041581
2.50676
8.60038
32.8486
3.32434
5.16779
9.14020
3.29124
0.002042
1.727
7.69253
18.4236
10.8390
69.3843
102.991
266.556
298.80
505.473
692.752
783.90
463.924
347.41
127.228
70.8994
13.8397
34
35
36
37
38
39
40
41
42
43
44
TOTALS
AVERAGE
3.70
4.10
4.10
4.00
4.50
4.60
4.50
4.60
4.60
4.70
5.00
36.00
31.10
31.70
38.50
47.90
49.10
55.80
70.10
70.90
79.10
94.00
32.66
38.0983
38.0983
36.74
43.5357
44.8951
43.5357
44.8951
44.8951
46.2544
50.3325
3.33918
11.15
6.99827 48.9757
6.39827 40.9378
1.76
3.10146
4.36428
19.05
4.20491 17.6813
12.2643
150.412
25.20
635.288
26.00
676.256
32.8456
1,078.83
43.6675
1,906.85
451.223
9,016.45
10.2551
204.92
(MAD)
(MSE)
Forecasting Summary Table

Method used:
Exponential
Smoothing
Linear Regression
(Trend Analysis)
Y = 18.968 +
1.638 YEAR
MAD
MSE
Standard error using
n 2 in denominator
Correlation coefficient
3.416
34.39
6.075
Linear Regression
10.587
171.817
13.416
Y = 17.636 +
13.59364
GSP
10.255
204.919
14.651
0.846
0.813
Giventhatonewishestodevelopafiveyearforecast,
trend analysis is the appropriate choice. Measures of
error and goodnessoffit are really irrelevant.
Exponential smoothing provides a forecast only of
depositsforthe next yearandthusdoesnotaddress
the fiveyear forecast problem. In order to use the
regression model based upon GSP, one must first
develop a model to forecast GSP, and then use the
forecastofGSPinthemodeltoforecastdeposits.This
requiresthedevelopmentoftwomodelsoneofwhich
(themodelforGSP)mustbebasedsolelyontimeasthe
independentvariable(timeistheonlyothervariablewe
aregiven).
(b)Onecouldmakeacaseforexclusionoftheolderdata.
Werewetoexcludedatafromroughlythefirst25years,
the forecasts for the later years would likely be
considerablymoreaccurate.Ourargumentwouldbethat
achangethatcausedanincreaseintherateofgrowth
appearstohavetakenplaceattheendofthatperiod.
Exclusionofthisdata,however,wouldnotchangeour
choice of forecasting model because we still need to
forecastdepositsforafuturefiveyearperiod.
ADDITIONAL HOMEWORK PROBLEMS

Theseproblems,whichappearonwww.myomlab.com,providean
additional13problemsthatyoumaywishtoassign.
4.50
(a)
(b)
(c)
Week
10
Forecast
Registration
Nave
2-week moving
4-week moving
22
21
22
25
21
21.5
27
25
23
35
27
26
23.7
5
29
35
31
27
33
29
32
29
37
33
31
31
41
37
35
33.5
37
41
39
35
37
39
37
51
52
4.51
(c)Based on
a Mean Absolute
Deviation
criterion, theAbsolute
Sales
Three-Month
Moving
Average
3month moving average
with MAD = 2.2 is to be
Moving
Deviation
Month
Period
1
2
3
4
5
6
7
4.52
Actual
95
108
123
130
Demand
Exponentially Smoothed
January
Forecast
February
7
5
March
9
5 + 0.2 (7 5) = 5.4
April
5
5.4 + 0.2 (9 5.4) = 6.12
May
9
6.12 + 0.2 (5 6.12) = 5.90
June
13
5.90 + 0.2 (9 5.90) = 6.52
July
8
6.52 + 0.2 (13 6.52) = 7.82
August
Forecast
7.82 + 0.2 (8 7.82) = 7.86
September
October
November
Forecast
|Error|
December
Error2
January
100
5
25
February
110
2
4
120
3
9
130
0
0
10
38
MAD=10/4=2.5,MSE=38/4=9.5
4.53(a)3monthmovingaverage:
Three-Month
Month
Sales
January
February
March
April
11
14
16
10
May
15
June
17
July
11
August
14
September
17
October
12
November
14
December
16
January
11
February
Moving Average
Absolut
e
Deviatio
n
preferredoverthe3monthweightedmovingaverage
11
withMAD=2.72.
14
(d)Otherfactorsthatmightbeincludedinamorecomplex
16
10 modelareinterestratesandcycleorseasonalfactors.
(1 11 + 2 14 + 3 16)/6 = 14.50
4.50
15
(1 14 + 2 16 + 3 10)/6 = 12.67
2.33
4.54(a)
17
(1 16 + 2 10 + 3 15)/6 = 13.50
3.50
11
(1 10 + 2Cumulativ
15 + 3 17)/6 = 15.17
Actua
Cum. Trackin 4.17
14l
(1 15 + 2 17e+ 3 11)/6 = 13.67
0.33
g
17
(1 17
+ 2 11
+ 3 14)/6
Wee Miles
Foreca
Error
Error
| = 13.50
MAD Signal 3.50
12
(1 11 + 2 14 + 3 17)/6
= 15.00
3.00
k
st
Error|
14
(1 14
+ 2 17 + 3 12)/6
0.00
1
17
17.00
0.00
0.00= 14.00
0
16
(1 17
+ 2 12
+ 3 14)/6
2.17
2
21
17.00
4.00
4.00
4.00= 13.83
2
2
11
(1 12
+ 2 14
+ 3 16)/6
3.67
3
19
17.80
1.20
5.20
5.20= 14.67
1.73 3
(1 14
+ 2 10.16
16 + 3 11)/6
4
23
18.04
4.96
10.16= 13.17
2.54 4
5
18
16
7
8
20
18
9
10
11
22
20
15
12
22
19.03 +1.0
3
18.83 +2.8
3
18.26 1.74
18.61 +0.6
1
18.49 3.51
19.19 0.81
19.35 +4.3
5
18.48 3.52
9.13
11.19
2.24 4
6.30
14.02
2.34
8.04
7.43
15.76
16.37
2.25
2.05
10.94
11.75
7.40
19.88
20.69
25.04
2.21 5
2.07
5.7
2.28
3.2
10.92
28.56
2.38
=27.17
MAD=2.72
2.7
3.6
3.6
4.6
(b)TheMAD=28.56/12=2.38
(11 + 14 + 16)/3
13.67
(14 + 16 + 10)/3
13.33
(16 + 10 + 15)/3
13.67
(10 + 15 + 17)/3
14.00
(15 + 17 + 11)/3
14.33
(17 + 11 + 14)/3
14.00
(11 + 14 + 17)/3
14.00
(14 + 17 + 12)/3
14.33
(17 + 12 + 14)/3
14.33
(12 + 14 + 16)/3
14.00
(14 + 16 + 11)/3
13.67
3.67
1.67
3.33
3.00
0.33
3.00
2.00
0.33
1.67
3.00
(c)The cumulativeerror andtrackingsignalsappearto

beconsistentlynegative,andatweek10,thetracking
signalexceeds5MADs.
4.55
x2
xy
7
9
5
11
10
13
55
1
2
3
4
5
6
21
1
4
9
16
25
36
91
7
18
15
44
50
78
212
y 9.17
x 3.5
y 5.27 1.11x
Period7forecast=13.07
Period12forecast=18.64,butthisisfaroutsidetherange
ofvaliddata.
=22.00
MAD=2.20
(b)3monthweightedmovingaverage
4.56Tocomputeseasonalizedoradjustedsalesforecast,wejust
multiplyeachseasonalizedindexbytheappropriatetrendforecast.
Y
Index Y
Seasonal
Hence,for
53
The correlation coefficient indicates that there is a positive

correlationbetweenbankdepositsandconsumerpriceindicesin
Trendforecast
QuarterI:YI 1.25 120,000 150,000

QuarterII:YII 0.90 140,000 126,000
QuarterIII:Y 0.75 160,000 120,000
III
QuarterIV:YIV 1.10 180,000 198,000

4.57
Mon
Week 1
Week 2
Week 3
Week 4
Averages
210
215
220
225
217.
5
Tue
Wed
178
180
176
178
17
8
250
250
260
260
255
Thu
215
213
220
225
218.
3
Fri
Sat
160
165
175
176
16
9
180
185
190
190
186.3
Overall average =
204
(a)Seasonalindices:
1.066 (Mon)0.873 (Tue)1.25 (Wed)
1.07 (Thu)0.828 (Fri)0.913 (Sat)
(b)To calculate for Monday of Week 5 = 201.74 +

0.18(25)1.066=219.9roundedto220
Forecast
Birmingham,Alabamai.e.,asonevariabletendstoincrease(or
decrease),theothertendstoincrease(ordecrease).
4.60(c)Standarderroroftheestimate:
220 (Mon) 180 (Tue) 258 (Wed)

221 (Thu) 171 (Fri)189 (Sat)
Syx
4.58(a)4000+0.20(15,000)=7,000
(b)4000+0.20(25,000)=9,000
4.59(a)35+20(80)+50(3.0)=1,785
(b)35+20(70)+50(2.5)=1,560
Y 2 a Y b XY
130 1 20 1 70
n2
3
130 20 70
4.61
4.60Given:X=15,Y=20,XY=70,X =55,Y =130,

X =3, Y =4
XY nXY
(a)
b
X 2 nX 2
a Y bX
70 5 3 4 70 60 10
b
1
55 45 10
55 5 32
a 4 1 3 4 3 1
Y 1 1X
Column Totals
X2
Y2
XY
2
1
4
5
3
4
1
4
6
5
4
1
16
25
9
16
1
16
36
25
8
1
16
30
15
15
20
55
94
70
Given:Y=a+bXwhere:
b
(b)Correlationcoefficient:
n X2 X
n Y 2 Y

5 70 15 20
XY nXY
X 2 nX 2
a Y bX
n XY X Y
40
13.3 3.65
3
5 55 152 5 130 20
2
350 300
50
50 250
275 225 650 400
50
0.45
111.80
andX=15, Y=20, XY=70, X =55, Y =94, X =3,

Y =4.Then:
b
70 5 4 3
2
55 5 3
a 4 1 3 1.0
70 60
1.0
55 45
andY=1.0+1.0X.Thecorrelationcoefficient:
r
n XY X Y
n X 2 X 2 n Y 2 Y 2
5 70 15 20
350 300
2
2
5 55 15 5 94 20
275 225 470 400
50
50
0.845
50 70 59.16
54
Standarderroroftheestimate:
Y 2 a Y b XY
n2
Syx
94 1 20 1 70
52
94 20 70
1.333 1.15
3
4.62Usingsoftware,theregressionequationis:Gameslost=
6.41+0.533daysrain.
CASE STUDIES
SOUTHWESTERNUNIVERSITY:B
Thisisthesecondinaseriesofintegratedcasestudiesthatrun
throughoutthetext.
1.Onewaytoaddressthecaseiswithseparateforecastingmodels
foreachgame.Clearly,thehomecominggame(week2)andthe
fourthgame(craftfestival)areuniqueattendancesituations.
Game
1
2
3
4
5
Total
Forecasts
2010
2011
Model
y
y
y
y
y
=
=
=
=
=
30,713
37,640
36,940
22,567
30,440
+
+
+
+
+
2,534x
2,146x
1,560x
2,143x
3,146x
48,453
52,660
47,860
37,567
52,460
239,00
0
50,988
54,806
49,420
39,710
55,606
250,53
0
R2
0.92
0.90
0.91
0.88
0.93
(wherey=attendanceandx=time)
2.Revenuein2010=(239,000)($50/ticket)=$11,950,000
Revenuein2011=(250,530)($52.50/ticket)=$13,152,825
3.In games 2 and 5, the forecast for 2011 exceeds stadium
capacity.Withthisappearingtobeacontinuingtrend,thetime
hascomeforaneworexpandedstadium.
2 DIGITALCELLPHONE,INC.
Objectives:
Totals
The importance of combining a qualitative model with a

quantitative modelinsituationswhere technologicalchangeis
occurring.
1
2
3
4
5
6
7
8
9
480
436
482
448
458
489
498
430
444
x2
1
4
9
16
25
36
49
64
81
xy
480
872
1,446
1,792
2,290
2,934
3,486
3,440
3,996
y2
230,400
190,096
232,324
200,704
209,464
239,121
248,004
184,900
197,136
666
19,366
16,206
537.9
4,960
5,357
6,300
7,475
7,378
8,100
8,032
8,636
10,314
9,652
9,960
10,185
11,572
12,696
14,088
15,200
15,522
16,524
16,884
18,212
18,150
19,437
18,496
19,305
19,754
22,120
23,616
246,016
237,169
275,625
330,625
277,729
291,600
252,004
258,064
328,329
258,064
248,004
235,225
276,676
304,704
344,569
369,664
356,409
374,544
363,609
394,384
366,025
393,129
334,084
342,225
337,561
399,424
430,336
378,661 10,558,246
450.2
10,518.4
293,284.6
xy nx y
100
121
144
169
196
225
256
289
324
361
400
441
484
529
576
625
676
729
784
841
900
961
1,024
1,089
1,156
1,225
1,296
378,661 36 18.5 537.9 20390.0
5.25
3885.0
x 2 nx 2
16,206 (36 18.52 )
a y bx 537.9 5.25 18.5 440.85
Selectionofanappropriatetimeseriesforecastingmodelbased
uponaplotofthedata.
496
487
525
575
527
540
502
508
573
508
498
485
526
552
587
608
597
612
603
628
605
627
578
585
581
632
656
Average 18.5
1.A plot of the data indicates a linear trend (least squares)

modelmightbeappropriateforforecasting.Usinglineartrendyou
obtainthefollowing:
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
n xy x y
2
[ n x ( x )2 ][ n y2 ( y)2 ]
(36)(378,661) (666)(19,366)
[(36) (16,206) (666)2 ][(36)(10,558,246) (19,366)2 ]
13,631,796 12,897,756
[(583,416) (443,556)][380,096,856) (375,041,956)]
737,040
[139,860][5,054,900]
734,040
706,978,314,000
734,040
.873
840,820
r 2 .76
y 440.85 5.25(time)
r=0.873indicatingareasonablygoodfit
Thestudentshouldreportthelineartrendresults,butdeflate
the forecast obtained based upon qualitative information about
industryandtechnologytrends.
Becausethereislimitedseasonalityinthedata,thelinear
trendanalysisaboveprovidesagoodr2of.76.
However,amorepreciseforecastcanbedevelopedaddressing
theseasonalityissue,whichisdonebelow.Methodsaandcyieldr2
of.85and.86,respectively,andmethodsbandd,whichalsocenter
theseasonaladjustment,yieldr2of.93and.94,respectively.
2.FourapproachestodecompositionofTheDigitalCellPhone
datacanaddressseasonality,asfollows:
10
Totals
a)Multiplicativeseasonalmodel,
Cases=443.87+5.08(time),r2=.85,MAD=20.89
Averag
e
b)MultiplicativeSeasonalModel,withcenteredmovingaverages
(CMA),whichisnotcoveredinourtextbutcanbeseenin
Render, Stair, and Hannas Quantitative Analysis for
Management,10thed.,PrenticeHallPublishing.
Cases=432.28+5.73(time),r2=.93,MAD=12.84
c)Additiveseasonalmodel,
Cases=444.29+5.06(time),r2=.86,MAD=20.02
d)Additiveseasonalmodel,withcenteredmovingaverages.
Cases=431.31+5.72(time),r2=.94,MAD=12.28
Thetwomethodsthatusetheaverageofalldatahavevery
similarresults,andthetwoCMAmethodsalsolookquiteclose.As
suggestedwiththisanalysis,CMAistypicallythebettertechnique.
60
36
6
36.6
66
37
6
37.6
55
3,600 4,356 3,960

15,950
15,77
15,910
2
15,772 10 36.6 37.6
0.7996 .8
15,910 10 36.62
a 37.6 0.7996 36.6 8.3363 8.3
Y 8.3363 0.7996 X
At$65,000;y8.3.8(65)8.352=60.3,or60,300guests.
Fortheinstructorwhoasksotherquestionsthanthisone:
r20.8869
Std.error5.062
ADDITIONAL CASE STUDY*

THENORTHSOUTHAIRLINE
VIDEO CASE STUDY
Northern Airline Data
FORECASTINGATHARDROCKCAFE
Thereisashortvideo(8minutes)availablefromPrenticeHall
andfilmedspecificallyforthistextthatsupplementsthiscase.
1.HardRockusesforecastingfor(1)sales(guestcounts)atcafes,
(2)retailsales,(3)banquetsales,(4)concertsales,(5)evaluating
managers,and(6)menuplanning.Theycouldalsoemploythese
techniquestoforecast:retailstoresalesofindividual(SKU)product
demands;salesofeachentre;salesateachworkstation,etc.
2.ThePOSsystemcapturesallthebasicsalesdataneededto
driveindividualcafesscheduling/ordering.Italsoisaggregated
atcorporateHQ.Eachentresoldiscountedasoneguestata
HardRockCafe.
3.Theweightingsystemissubjective,butisreasonable.More
weightisgiventoeachofthepast2yearsthanto3yearsago.
Thissystemactuallyprotectsmanagersfromlargesalesvariations
outside their control. One could also justify a 50%30%20%
modelorsomeothervariation.
4.Otherpredictorsofcafesalescouldincludeseasonofyear
(weather); hotel occupancy; spring break from colleges; beef
prices;promotionalbudget;etc.
5.Yabx
Month Advertising Guest Count
X
Y
1
2
3
4
5
14
17
25
25
35
21
24
27
32
29
35
37
45
43
50
43
60
54
X2
Y2
XY
196
289
625
625
1,225
1,225
2,025
2,500
3,600
441
576
729
1,024
841
294
408
675
800
1,015
1,369
1,295
1,849
1,935
1,849
2,150
2,916
3,240
Year
Airframe Cost
per Aircraft
Engine Cost
per Aircraft
Average
Age (hrs)
2003
2004
2005
2006
2007
2008
2009
51.80
54.92
69.70
68.90
63.72
84.73
78.74
43.49
38.58
51.48
58.72
45.47
50.26
79.60
6512
8404
11077
11717
13275
15215
18390
Year
Airframe Cost
per Aircraft
Engine Cost
per Aircraft
Average
Age (hrs)
2003
2004
2005
2006
2007
2008
2009
13.29
25.15
32.18
31.78
25.34
32.78
35.56
18.86
31.55
40.43
22.10
19.69
32.58
38.07
5107
8145
7360
5773
7150
9364
8259
Southeast Airline Data
Utilizingthesoftwarepackageprovidedwiththistext,we
candevelopthefollowingregressionequationsforthevariables
ofinterest:
NorthernAirlinesAirframeMaintenanceCost:
Cost=36.10+0.0026Airframeage
Coefficientofdetermination=0.7695
Coefficientofcorrelation=0.8772
NorthernAirlinesEngineMaintenanceCost:
SoutheastAirlinesAirframeMaintenanceCost:
56
SoutheastAirlinesEngineMaintenanceCost;
Thefollowinggraphsportrayboththeactualdataandthe
regression lines for airframe and engine maintenance costs for
bothairlines.
Overall,itwouldseemthat:
*ThiscasestudyappearsonourcompanionWebsite,at
www.pearsonhighered.com/heizer,andatwww.myomlab.com.
Southeast Airlines: The relationships between mainte

nance costs and airframe age for Southeast Airlines are
much less well defined. It is even more obvious that
airframeageisnottheonlyimportantfactorperhapsnot
eventhemostimportantfactor.
Northern Airlines has the smallest variance in mainte

nancecostsindicatingthatitsdaytodaymanagement
ofmaintenanceisworkingprettywell.
Maintenancecostsseemtobemoreafunctionofairline
thanofairframeage.
TheairframeandenginemaintenancecostsforSoutheast
Airlinesarenotonlylower,butmorenearlysimilarthan
thoseforNorthernAirlines.Fromthegraphs,atleast,they
appeartoberisingmoresharplywithage.
Fromanoverallperspective,itappearsthatSoutheastAirlines
may perform more efficiently on sporadic or emergency
repairs,andNorthernAirlinesmayplacemoreemphasis
onpreventivemaintenance.
Ms.Youngsreportshouldconcludethat:
Thereisevidencetosuggestthatmaintenancecostscould
bemadetobeafunctionofairframeagebyimplementing
moreeffectivemanagementpractices.
Thedifferencebetweenmaintenanceproceduresofthetwo
airlinesshouldbeinvestigated.
The data with which she is currently working does not
provideconclusiveresults.
ConcludingComment:
Thequestionalwaysarises,withthiscase,astowhetherthedata
shouldbemergedforthetwoairlines,resultingintworegressions
insteadoffour. Thesolutionprovidedisthatoftheconsultant
who was hired to analyze the data. The airlines own internal
analystsalsoconductedregressions,but did mergethedatasets.
Thisshowshowstatisticianscantakedifferentviewsofthesame
data.
Notethatthetwographshavebeendrawntothesamescale
tofacilitatecomparisonsbetweenthetwoairlines.
Comparison:
NorthernAirlines:Thereseemtobemodestcorrelations
betweenmaintenancecostsandairframeageforNorthern
Airlines.Thereiscertainlyreasontoconclude,however,
thatairframeageisnottheonlyimportantfactor.

Heizerom10ism04 141023093306 Conversion Gate02

Transféré par

Informations du document

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Heizerom10ism04 141023093306 Conversion Gate02

Transféré par

Droits d'auteur :

Formats disponibles

CHAPTER 4F O R E C A S T I N G

Active Model Exercises*

13.0 9.0 11.0 7.0

(d)The threeyear moving average appears to give better

1.Scroll through different values for alpha and beta. Which

9.0 13.0 8.0 12.0 13.0 9.0 11.0 7.0

TEACHING NOTE: Notice how well exponential

4.4 (a)FJuly FJune 0.2(Forecastingerror)

(b)FAugust FJuly 0.2(Forecastingerror)

42 0.2(40 42) 41.6

5.0 10.0 8.0

5.0 10.0 8.0

Then:X=15,Y=261,XY=815,X =55, X =3, Y =52.2

Exponential smoothing, = 0.6

(450 + 495 + 518)/3 =

8166 (5)(3)(522) 336

a Y bX 522 (33.6)(3) 421.2

0.6(450 410) = 434.0

0.9(450 410) = 446.0

MAD = 24.3/6 = 4.05, MSE = 113.2/6 = 18.87. Note that all

3.8 16.11 19.91

4.8 17.71 22.51

1.04 1.34 2.38

FIT6 F6 T6 22.51 2.38 24.89

0.6 2.07 2.45

FIT8 F8 T8 27.14 2.45 29.59

0.8 29.59 29.28

T9 F9 F8 1 T8 0.4 29.28 27.14

0.6 2.45 2.32

FIT9 F9 T9 29.28 2.32 31.60

Actual Nave |Error| |% Error|

(d) R=.82isthecorrelationcoefficient,andR 2 =.68

X =33, Y =38, XY =227, X2 =199, X =5.5, Y =6.33.

(c) If there are nine performances by Stone Temple

Yr1 Demand Yr2 Demand

SumofAveYr1 to Yr2 Demand

Quarte 2007 2008

(9, 783) 6(3.25833)(550)

Cumulative error = 14.05; MAD = 5Tracking = 14.05/5

factors besides square footage that will impact the

(d)Coefficient of determination = (0.63)2 = 0.397. This

Number of days on the road X1 = 5 and distance

2900 10 5.5 47.8

Error Running sum |error|

2900 2629 271

4.other transportation costscab, limousine, special

X 11:Y 29.76 3.28 11 65.8

Perio Demand Forecas

andX=55, Y=478, XY=2900, X 2 =385, Y 2 =23892,

and X = 854, Y = 478, XY = 42558.6, X2 = 76129.9,

42558.6 10 85.4 47.8

The MAD is 3.26this is approximately 7% of the average

X 131.2 : Y 1.43 0.543(131.2) 72.7

and X = 132, Y = 27.1, XY = 352.9, X2 = 1796,

71.59 13.85 56.11

12 352.9 132 27.1

12 1796 1322 12 71.59 27.1

To evaluate the trend adjusted exponential smoothing model,

and X =6885, Y =36.96, XY =20299.5, X 2 =3857893,

20299.5 13 529.6 2.843

Amits MAD for exponential smoothing (16.11) is lower

13 20299.5 6885 36.96

Barba Amit Error Barbara

(d) Note that Amit has more forecast observations, while