Chapter 04

24
CHAPTER 4F O R E C A S T I N G
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(under 3months), mediumrange (3months to
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
model incorporatesinto themodel thevariables offactors that
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 indexes 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
befilteredoutforgoodmediumrangeplanning(ofproduction
andinventory)andperformanceevaluation.
20.Therearemanyexamples.Demandforrawmaterialsand
componentpartssuchassteelortiresisafunctionofdemandfor
goodssuchasautomobiles.
25
21.Obviously,aswegofartherintothefuture,itbecomesmore
difficulttomakeforecasts,andwemustdiminishourrelianceon
theforecasts.
*ActiveModels4.1,4.3,and4.4appearontheCDROMonly.
ActiveModel4.2alsoappearsinthetext.
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(anaiveforecast)
ACTIVEMODEL4.2:ExponentialSmoothing
1.Whathappenstothegraphwhenalphaequalszero?
Thegraphisastraightline.Theforecastisthesamein
eachperiod.
2.Whathappenstothegraphwhenalphaequalsone?
The forecast follows the same pattern as the demand
(exceptforthefirstforecast)butisoffsetbyoneperiod.This
isanaiveforecast.
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
26
(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.Withbeta setto zero, findthe bestalpha andobserve the
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?
Notheyarenotthesamevalues.Forexample,anintercept
of57.81withaslopeof9.44yieldsaMADof7.17.
END-OF-CHAPTER PROBLEMS
4.1(a)
TEACHING NOTE: Notice how well exponential

smoothingforecaststhenaive.
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
Naivetrackstheupsanddownsbestbutlagsthedatabyone
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
August 31
September
September
September
September
October 5
4.5(a)
360
389
410
381
368
Forecasti Error
ng
.20
Error
0
0
360
360
29
5.8
365.8
44.2
8.84
374.64
6.36
1.272
375.912 7.912
1.5824
374
374.3296.3296
Pints Forecas
t
7
14
21
28
Foreca
st
360
365.8
374.64
375.912
374.329
6
.06592 374.263
6
3,700 + 3,800
=3,750miles
2
(b)
Year
Mileage
Two-Year
Moving Average
Error
|Error|
27
1
2
3
4
5
3,000
4,000
3,400
3,800
3,700
4.5(c)Weighted2yearM.A.with.6weightformostrecentyear.
3,500
3,700
3,600
Totals
MAD
100
100
100
10
0
100
100
100
300
Year
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
42
0
1
2
3
4
5
300
100.
3
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
3,000
3,000
3,500
3,450
3,625
Total
Forecast
Error
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
(a)
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
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
1474
(b)[i]Naive
ThecomingJanuary=December=23
[ii]3monthmoving(20+21+23)/3=21.33
(b)
28
(88 90)
89
2
(c)
[iii]6monthweighted[(0.117)+(.118)
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
+(0.120)+(0.220)
+(0.221)+(0.323)]/1.0
=20.6
[iv]Exponentialsmoothingwithalpha=0.3
Moving
Average
1.735
1.685
1.775
1.875
1.885
1.835
1.815
1.765
1.675
1.675
93
94
93
0.5 0.25 100(.5/93)
93.5
FNov 18.6 0.3(20 18.6) 19.02
96
94.0
88
95.5
90
92.0
1474 (12)(6.5)(18.2) 54.4
0.38
650 12(6.5)2
143
a 18.2 0.38(6.5) 15.73
Forecast=15.73+.38(13)=20.67,wherenextJanuary
isthe13thmonth.
(c)Onlytrendprovidesanequationthatcanextendbeyond
onemonth
4.7Present=Period(week)6.
1
1
1 1
So: F7 3 A6 4 A5 4 A4 6 A3

(52) +
3

1
1
1
(63) + (48) + (70) = 56.75patients
4
4
6
1 1 1 1
where 1.0 = weights , , ,
3 4 4 6
4.8(a)
1.0
(96 88 90)
91.3
3
=
0.54%
93.5
[v]Trend x 78,x 6.5, y = 218,y 18.17
.035
.165
.125
.005
.085
.005
.115
.115
.025
.075
.
Absolute % Error
Temperatur 2 day
e
M.A.
95
FJan 19.6 0.3(23 19.6) 20.62 21
|Error|
Two-Month
ThreeMonth
Moving
Moving
Average
Average
1.723
1.740
1.817
1.873
1.857
1.833
1.777
1.727
1.683
Totals
|
(Error
Error|
)2
FOct 18 0.3(20 18) 18.6

FDec 19.02 0.3(21 19.02) 19.6
ThreeMonth
Moving
Average
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.
66.7
14.94
5
5
%
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.
.127
.160
.053
.073
.027
.133
.127
.027
.067
.793
29
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)MAD(twomonthmovingaverage)=.750/10=.075
MAD(threemonthmovingaverage)=.793/9=.088
Therefore, the twomonth moving average seems to have
performedbetter.
(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
30
(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=0.5better
MSE:0.04+0.0025+0.0025+0.04=0.085
Method2:
MAD:0.1+0.20+0.10+0.11=0.51
MSE:0.01+0.04+0.01+0.0121=0.072better
31
4.15
Year
Sales
2003
2004
2005
2006
450
495
518
563
2007
584
2008
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
2003
2004
2005
2006
2007
1
2
3
4
5
450
495
518
563
584
1
4
9
16
25
450
990
1554
2252
2920
= 55
=
8166
2610
XY
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
2003
2004
2005
2006
2007
2008
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
2003
2004
2005
2006
450
495
518
563
2007
2008
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
Forecast Exponential
Year
Sales
2003
2004
2005
2006
450
495
518
563
2007
584
2008
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
+ 0.9(584 558.2) =
=190.5
MAD=38.1
(RefertoSolvedProblem4.1)
For =0.3,absolutedeviationsfor20032007are40.0,73.0,
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 8.4
1.2 1.6
5.6 31.9
4.4 19.7
5.2 26.6
24.3 113.
2
MAD = 24.3/6 = 4.05, MSE = 113.2/6 = 18.87. Note that all

numbersarerounded.
32
Note:TousePOMforWindowstosolvethisproblem,aperiod0,
whichcontainstheinitialforecastandinitialtrend,mustbeadded.
33
4.20Trendadjustedexponentialsmoothing:=0.1,=0.8
4.23Students must determine the naive 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.ThenaiveforecastforMarchistheFebruaryactualof83,
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)
Note: To use POM for

Windows to solve this
problem, a period 0, which
contains the initial forecast
and initial trend, must be
added.
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.ItsMSEof12.70islower.ItsMADof3.39is
T6 F6 F5 1 T5 0.4
22.51
19.91
0.6 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
F8 A7 (1 )( F7 T7 ) (0.2)(31)
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(formanager)
15.29%
4
MAD(formanager)
(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 Naive |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
%
(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(fornaive)
12.12%.
4
Naiveoutperformsmanagement.
MAD(fornaive)
4.25
(c)MADforthemanagerstechniqueis14.5,whileMADforthe
naiveforecastisonly12.25.MAPEsare15.29%and12.12%,
respectively.Sothenaivemethodisbetter.
4.24(a)Graphofdemand
Theobservationsobviouslydonotformastraightlinebutdotend
toclusteraboutastraightlineovertherangeshown.
(b)Leastsquaresregression:
Month
Number of
Accidents
(y)
January
February
March
April
Totals
Averages
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:
IftherearenineperformancesbyStoneTemplePilots,the
estimatedsalesare:
Y9 .676 1.03 9 .676 9.27 9.93 drums
10drums
34
20,000
1, 250
16
6,000
Averageoverspring:
1,500
4
1,500
Springindex:
1.2
1, 250
Averageoverallseasons:
5,600
(1.2) 1,680sailboats
4
Answer:
xy
x2
30
80
180
360
650
1
4
9
16
30
35
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
2004
2005
2006
2007
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 2005 2006

r
Winter
Spring
Summer
Fall
73
104
168
74
65
82
124
52
2007
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
4.32(a)
x2
xy
16
330
5,280
12
270
3,240
18
380
6,840
14
300
CHAPTER
4F
O R E C A S4,200
TING
60
1,280
19,560
4.292009is25yearsbeyond1984.Therefore,the2009quarter
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
Soatx=1.80,y=1,454.6277.6($1.80)=954.90.Nowroundto
thenearestinteger:Answer:955lattes.
(5)
(4)
Adjusted
Seasonal Forecast
Factor
[(3) (4)]
.8
1.1
1.4
.7
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
x
96.344
132.946
169.806
85.204
4.30GivenY=36+4.3X
(a) Y=36+4.3(70)=337
(b)Iftheforecastisfor20guests,thebarsalesforecast
is 50 + 18(20) = $410. Each guest accounts for an
additional$18inbarsales.
(b)Y=36+4.3(80)=380
(c) Y=36+4.3(90)=423
4.31
Letx1, x2 , K , x6 bethepricesandy1, y2 , K ,y6

bethenumbersold.
(b)MSE=160/5=32
(c)MAPE=13.23%/5=2.65%
i 1
6
Y
6
4.34Y=7.5+3.5X1+4.5X2+2.5X3
xi
Averageprice=3.2583
yi
i 1
256
144
324
196 36
920
(a)28
(b)43
(1)
4.33(a)Seethetablebelow.
Averagenumbersold=550
(2)
2,880 5(3)(180) 2,880 2,700
(3)
b
55 45
55 5(3)2
180
(4)
18
10
a 180 3(18) 180 54 126
y 126 18 x
xi yi 9,783
i 1
6
2
xi =67.1925
i1
TableforProblem4.33
Year
(x)
Then y = a + bx,
6
xi yi nx y
i 1
6
i 1
x2i
nx
Transistors
(y)
xy
x2
126 + 18x
Error
Error 2
140
140
144
16
160
320
162
(9, 783) 6(3.25833)(550)

3
190
570
180
10
100
200
800
16
198
210
1,050
25
216
36
2,80
0
5
5
67.1925 6(3.25833)2
969.489
277.6
3.49222
a y bx 1, 454.6
Totals
1
5
x = 3
where y = number
sold,x=price,and
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
37
(b)Thepredictedsellingpriceis$83,502,butthisisthe
averagepriceforahouseofthissize.Thereareother
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.
(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
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
Year X
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
1
2
3
4
5
6
7
8
9
10
55
Given:Y=a+bXwhere:
r 0.68(coefficientofcorrelation)
If:
Number of days on the road X1 = 5 and distance
traveledX2=300
then:
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:
3.costsofentertainingcustomers
andX =55, Y =478, XY =2900, X 2 =385, Y 2 =23892,

X 5.5,Y 47.8, Then:
b
2900 10 5.5 47.8

2
4.37(a,b)
Perio Demand Forecas
d
t
1
2
3
4
5
6
7
8
9
10
20
21
28
37
25
29
36
22
25
28
20
20
20.5
24.25
30.63
27.81
28.41
32.20
27.11
26.05
RSFE = 14.05; MAD = 5
Error Running sum |error|

0.00
1.00
7.50
12.75
5.63
1.19
7.59
10.20
2.10
1.95
0.00
1.00
8.50
21.25
15.63
16.82
24.41
14.21
12.10
14.05
0.00
1.00
7.50
12.75
5.63
1.19
7.59
10.20
2.10
1.95
MAD 5.0
0
Tracking = 14.05/5 2.82
385 10 5.5
a 47.8 3.28 5.5 29.76
2900 2629 271
3.28
385 302.5 82.5
andY=29.76+3.28X.For:
X 11:Y 29.76 3.28 11 65.8
X 12:Y 29.76 3.28 12 69.1
Therefore:
Year1165.8patients
4.other transportation costscab, limousine, special

tolls,orparking
Inaddition,thecorrelationcoefficientof0.68isnotexceptionally
high.Itindicatesthatthemodelexplainsapproximately46%of
the overall variation in trip cost. This correlation coefficient
wouldsuggestthatthemodelisnotaparticularlygoodone.
X 2 nX 2
a Y bX
1.thetypeoftravel(airorcar)
2.conferencefees,ifany
XY nXY
Year1269.1patients
The model seems to fit the data pretty well. One should,
however,bemorepreciseinjudgingtheadequacyofthemodel.
Twopossibleapproachesarecomputationof(a)thecorrelation
coefficient, or (b) the mean absolute deviation. Thecorrelation
coefficient:
r
n XY X Y
n X X 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
r 2 0.853
2710
0.924
2934.3
Thecoefficientofdeterminationof0.853isquiterespectable
indicatingouroriginaljudgmentofagoodfitwasappropriate.
38
39
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
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.
X 2 nX 2
a Y bX
1.9
=32.6
MAD=3.26
XY nXY
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
scatteraroundastraightline.
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
and X = 132, Y = 27.1, XY = 352.9, X2 = 1796,

Y 2 =71.59, X =11, Y =2.26.Then:
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)
40
41
(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.
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,anaivemodelthat
studentscreateontheirownmightbebest.
(b)Onemodelmightbe:Ft+1=At11
Thatisforecast nextperiod=actualoneyearearliertoaccount
for
seasonality.Butthisignoresthetrend.
Oneverygoodapproachwouldbetocalculatetheincrease
fromeachmonthlastyeartoeachmonththisyear,sumall12
increases, and divide by 12. The forecast for next year would
equal thevalue forthe samemonth thisyear plusthe average
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.
42
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
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
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.
Tracking signal
Y2
377 2.93
142129 8.58
585 3.00
342225 9.00
690 3.45
476100 11.90
( At Ft )
608 3.66
369664 13.40
MAD
390 2.88
152100 8.29
t 1
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
X2
177241 8.41
Month
So:MAD:
421 2.90
4.45
Note:TousePOMforWindows
tosolvethisproblem,aperiod0,
which contains the initial fore
cast and initial trend, must be
added.
|At Ft |
0
24
11
14
(At Ft )
0
24
11
14
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
20
20
11
11
Sum: 87
Sum: 39
Column totals
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
43
XY nXY
X 2 nX 2
a Y bX
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
andY=1.03+0.0034X
Asanindicationoftheusefulnessofthisrelationship,wecan
calculatethecorrelationcoefficient:
r
n XY X Y
n X 2 X 2
2
n Y 2 Y
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
X 350:Y 1.03 0.0034 350 2.22
X 800:Y 1.03 0.0034 800 3.75
Note:Whensolvingthisproblem,caremustbetakentointerpret
significantdigits.
4.47(a)Thereisnotastronglineartrendinsalesovertime.
(b, c)Amit wants to forecast by exponential smoothing
(setting Februarys forecast equal to Januarys sales)
with alpha 0.1. Barbara wants to use a 3period
moving
average.
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
notethattheMADforAmitisanaverageof4numbers,
whileBarbarasisonly2.
44
Amits MAD for exponential smoothing (16.11) is lower

than that of Barbaras moving average (19.17). So his
forecastseemstobebetter.
45
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
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
4.49(a)(Continued)
Year
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
TOTALS
AVERAGE
Method Exponential Smoothing

0.6 =
Deposits Forecast
|Error|
Error2
(Y )
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
38.50
47.90
49.10
55.80
70.10
70.90
79.10
94.00
787.30
17.8932
31.71
35.784
43.0536
46.6814
52.1526
62.9210
67.7084
74.5434
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
1,513.22
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
30
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
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
18.10
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
30.18
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
145.936
31
32
33
34
35
36
37
38
39
40
41
42
43
44
31
32
33
34
35
36
37
38
39
40
41
42
43
44
TOTALS
990.0
0
AVERAGE
22.50
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
787.3
0
17.893
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
46
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
7,559.9
5
171.817
(MSE)
Method Least squaresSimple Regression on

GSP
a
b
17.636 13.5936
Coefficients GPS 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
34
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
3.70
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
36.00
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
32.66
3.33918
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
11.15
47
35
36
37
38
39
40
41
42
43
44
TOTALS
AVERAGE
4.10
4.10
4.00
4.50
4.60
4.50
4.60
4.60
4.70
5.00
31.10
31.70
38.50
47.90
49.10
55.80
70.10
70.90
79.10
94.00
38.0983
38.0983
36.74
43.5357
44.8951
43.5357
44.8951
44.8951
46.2544
50.3325
6.99827
6.39827
1.76
4.36428
4.20491
12.2643
25.20
26.00
32.8456
43.6675
451.223
10.2551
(MAD)
48.9757
40.9378
3.10146
19.05
17.6813
150.412
635.288
676.256
1,078.83
1,906.85
9,016.45
204.92
(MSE)
2
3
4
5
Total
y
y
y
y
=
=
=
=
37,640
36,940
22,567
30,440
+
+
+
+
2,146x
1,560x
2,143x
3,146x
52,660
47,860
37,567
52,460
239,00
0
54,806
49,420
39,710
55,606
250,53
0
0.90
0.91
0.88
0.93
(wherey=attendanceandx=time)
2.Revenuein2008=(239,000)($20/ticket)=$4,780,000
Revenuein2009=(250,530)($21/ticket)=$5,261,130
3.In games 2 and 5, the forecast for 2009 exceeds stadium
capacity.Withthisappearingtobeacontinuingtrend,thetime
Forecasting Summary Table

Exponential
Smoothing
Method used:
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
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.
CASE STUDIES
1
SOUTHWESTERNUNIVERSITY:B
Thisisthesecondinaseriesofintegratedcasestudiesthatrun
throughoutthetext.
1.Onewaytoaddressthecaseiswithseparateforecastingmodels
foreachgame.Clearly,thehomecominggame(week2)andthe
fourthgame(craftfestival)areuniqueattendancesituations.
Game
1
Model
y = 30,713 + 2,534x
Forecasts
2008
2009
48,453
50,988
Linear Regression
R2
0.92
hascomeforaneworexpandedstadium.
2 DIGITALCELLPHONE,INC.
Objectives:
Selectionofanappropriatetimeseriesforecastingmodelbased
uponaplotofthedata.
The importance of combining a qualitative model with a
quantitativemodel in situations wheretechnological change is
occurring.
1.A plot of the data indicates a linear trend (least squares)

modelmightbeappropriateforforecasting.Usinglineartrendyou
obtainthefollowing:
x
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
y
480
436
482
448
458
489
498
430
444
496
487
525
575
527
540
502
508
573
508
498
x2
1
4
9
16
25
36
49
64
81
100
121
144
169
196
225
256
289
324
361
400
xy
480
872
1,446
1,792
2,290
2,934
3,486
3,440
3,996
4,960
5,357
6,300
7,475
7,378
8,100
8,032
8,636
10,314
9,652
9,960
y2
230,400
190,096
232,324
200,704
209,464
239,121
248,004
184,900
197,136
246,016
237,169
275,625
330,625
277,729
291,600
252,004
258,064
328,329
258,064
248,004
Totals
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
485
526
552
587
608
597
612
603
628
605
627
578
585
581
632
656
441
484
529
576
625
676
729
784
841
900
961
1,024
1,089
1,156
1,225
1,296
666
19,366
16,206
Average 18.5
537.9
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
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
378,661 36 18.5 537.9 20390.0
5.25
3885.0
S x 2 nx 2
16,206 (36 18.52 )
a y bx 537.9 5.25 18.5 440.85
S xy nx y
nS xy S xS y
2
[ nS x ( S x )2 ][ nS y 2 (S 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:
a)Multiplicativeseasonalmodel,
Cases=443.87+5.08(time),r2=.85,MAD=20.89
b)MultiplicativeSeasonalModel,withcenteredmovingaverages
(CMA),whichisnotcoveredinourtextbutcanbeseenin
Render, Stair, and Hannas Quantitative Analysis for
Management,9thed.,PrenticeHallPublishing.
Cases=432.28+5.73(time),r2=.93,MAD=12.84
c)Additiveseasonalmodel,
48
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.
VIDEO CASE STUDY

FORECASTINGATHARDROCKCAFE
Thereisashortvideo(8minutes)availablefromPrenticeHall
andfilmedspecificallyforthistextthatsupplementsthiscase.
49
A2minuteversionofthevideoalsoappearsonthestudentDVD
inthetext.
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
10
Totals
60
36
6
36.6
Averag
e
66
37
6
37.6
X2
Y2
XY
196
441
294
289
576
408
625
729
675
625 1,024
800
841 1,015
1,225
1,369 1,295
1,225
1,849 1,935
2,025
1,849
2,150
2,500
2,916
3,240
3,600
3,600 4,356 3,960
15,91015,950 15,77
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
50
*ThiscasestudyappearsonourcompanionWebsite,atwww.prenhall.
com/heizer.
INTERNET CASE STUDY*

THENORTHSOUTHAIRLINE
Northern Airline Data
Year
Airframe Cost
per Aircraft
Engine Cost
per Aircraft
Average
Age (hrs)
2001
2002
2003
2004
2005
2006
2007
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
Southeast Airline Data

Year
Airframe Cost
per Aircraft
Engine Cost
per Aircraft
Average
Age (hrs)
2001
2002
2003
2004
2005
2006
2007
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
Utilizingthesoftwarepackageprovidedwiththistext,we
candevelopthefollowingregressionequationsforthevariables
ofinterest:
NorthernAirlinesAirframeMaintenanceCost:
Cost=36.10+0.0026Airframeage
Coefficientofdetermination=0.7695
Coefficientofcorrelation=0.8772
NorthernAirlinesEngineMaintenanceCost:
SoutheastAirlinesAirframeMaintenanceCost:
SoutheastAirlinesEngineMaintenanceCost;
Thefollowinggraphsportrayboththeactualdataandthe
regression lines for airframe and engine maintenance costs for
bothairlines.
Notethatthetwographshavebeendrawntothesamescale
tofacilitatecomparisonsbetweenthetwoairlines.
Comparison:
NorthernAirlines:Thereseemtobemodestcorrelations
betweenmaintenancecostsandairframeageforNorthern
Airlines.Thereiscertainlyreasontoconclude,however,
thatairframeageisnottheonlyimportantfactor.
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.
Overall,itwouldseemthat:
Northern Airlines has the smallest variance in mainte

nancecostsindicatingthatitsdaytodaymanagement
ofmaintenanceisworkingprettywell.
Maintenancecostsseemtobemoreafunctionofairline
thanofairframeage.
TheairframeandenginemaintenancecostsforSoutheast
Airlinesarenotonlylower,butmorenearlysimilarthan
51
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.

Chapter 04

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Chapter 04

Caricato da

Copyright:

Formati disponibili

24

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

0.5 0.25 100(.5/93)

FNov 18.6 0.3(20 18.6) 19.02

1474 (12)(6.5)(18.2) 54.4

[v]Trend x 78,x 6.5, y = 218,y 18.17

FJan 19.6 0.3(23 19.6) 20.62 21

FOct 18 0.3(20 18) 18.6

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

Note: To use POM for

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 Naive |Error| |% Error|

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

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

Yr1 Demand Yr2 Demand

SumofAveYr1 to Yr2 Demand

Quarte 2005 2006

Letx1, x2 , K , x6 bethepricesandy1, y2 , K ,y6

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

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

2900 10 5.5 47.8

RSFE = 14.05; MAD = 5

Error Running sum |error|

2900 2629 271

4.other transportation costscab, limousine, special

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

13 20299.5 6885 36.96

Barba Amit Error Barbara