spss Assignment of Data Analysis.docx
《spss Assignment of Data Analysis.docx》由会员分享,可在线阅读,更多相关《spss Assignment of Data Analysis.docx(13页珍藏版)》请在冰豆网上搜索。
spssAssignmentofDataAnalysis
UniversityofGlamorgan
MScManagementandDevelopmentofInternationalFinancialSystems
Module:
TechniquesofDataAnalysisforFinancialMarkets
Assignment:
StatisticalStudy
To:
UniversityofGlamorgan
From:
StudentNumberXXXXXXXX
Date:
XXXXXXXXXX
StatisticalResearchStudy:
AnalysingtheRelationshipbetweenEPSandGearingeitherinLeisureorPharmaceuticalIndustry
Content
1.Introduction----------------------------------------------------------------------------3
2.TheRelationshipBetweenEPSandGearing-------------------------------3
3.PopulationandRandomSampling---------------------------------------------4
4.HistogramsandTestofNormality---------------------------------------------7
4.1Histograms-------------------------------------------------------------------------7
4.2TestofNormality----------------------------------------------------------------8
5.CorrelationAnalysis------------------------------------------------------------------9
5.1Formulatehypotheses-----------------------------------------------------------9
5.2ScatterPlots-----------------------------------------------------------------------10
5.3Spearman’sRankOrderCorrelation--------------------------------------11
6.DifferencesAnalysis-----------------------------------------------------------------12
6.1Formulatehypotheses-----------------------------------------------------------12
6.2TestofNormalityforDifferencesAnalysis-----------------------------12
6.3Mann-WhitneyUTest----------------------------------------------------------13
7.Limitations--------------------------------------------------------------------------------13
8.Conclusion--------------------------------------------------------------------------------14
References-----------------------------------------------------------------------------------15
1.Introduction
Itiswidelyacceptedthatearningpershare(EPS)isimportanttothecorporatefinancialdecisions(DavidsonandMallin,1998)andanychangesintheuseofdebtwouldcausechangesinEPS(Dickersonetal.,1995).ThepurposeofthisreportistofindoutifthereisarelationshipbetweenEPSandGearingeitherinleisureorpharmaceuticalindustrythroughaseriesoftestsbySPSSandifthereis,whatleveloftherelationshipis.
Thestudyusesaquantitativeapproach.Firstly,itwillcollectdataof80companiesfromtwoindustries“Leisure”and“Pharmaceutical”whicharealllistedonLondonStockExchangeandthenget30fromeachindustryasrandomsamplesforthetests.Secondly,itwouldassessthenormalityforthesamplesbothbyhistogramsandthe“Explore”optionofSPSSandthirdlychoosetheappropriatetestofcorrelationtoworkoutifthereisarelationshipbetweenthesetwofinancialvariablesornot.Finally,itwouldtestthedifferencesbetweenthetwosectorsmeansofEPSandGearing.
Themethodologyusedtoanswertheresearchquestionswastoselectanappropriatestatisticalpackage(SPSSused),runappropriatestatisticaltestsandinterpretresultsinaclearandconciseformat.
2.TheRelationshipBetweenEPSandGearing
Gearingoccurwhenabusinessisfinanced,atleastinpart,bycontributionsfromoutsideparties.Thelevelofgearingassociatedwithabusinessisoftenanimportantfactorinassessingrisk(AtrillandMcLaney,2001).EPSistheportionofacompany'sprofitallocatedtoeachoutstandingshareofcommonstock.Itservesasanindicatorofacompany'sprofitability.Earningpershareisgenerallyconsideredtobethesinglemost importantvariableindeterminingashare'sprice.SuchasEPSandGearing,thesefinancialvariablesareveryimportantincorporatefinancialdecisions.Forexample,capitalchoicesarebasedonthetrade-offbetweenEPSandthecapitalgearingofthefirm(DavidsonandMallin,1998).Naturally,thehighergearingwillbringboththeriskierdebtandhigherexpectedearnings(Dickersonetal.,1995).However,theremaybedifferentsituationsamongthedifferentindustriesandadditionallyexcessivelevelsofgearingareoftenresponsibleforcorporatefailure(PikeandNeale,1996).Thereby,itisverynecessarytodoresearchontherelationshipbetweenEPSandGearingespeciallywhenfirmsneedtoconsiderthemasimportantelementsforcapitaldecisions.
3.PopulationandRandomSampling
Apopulationisthetotalcollectionofelementswhichisusedduringthetests(Blumberg,2005).Inthisreport,allthe80selectedcompaniesarelistedontheLondonStockExchangeand40ofthosecompaniesarefrom“Leisure”industrywhiletheothersarefrom“Pharmaceutical”withthedataoftwofinancialvariablesEPSandGearing,additionally,ithasbeencheckedthattherearenoerrors(e.g.missing)indatafiles.
Usually,itisassumedthattheselectionreflectsthecharacteristicsofthefullset(Blumbergetal.,2005)andthereforetherandomsamplingistheinfluentialstepofthewholeprocessandtheeveryelementsofthepopulationshouldhaveaknown,non-zeroprobabilityofbeingincludedinthesample(Gillbert,D.etal,2004).Accordingly,30companiesarerandomlychosenfromeachsectorbySPSSandhasbeenarrangedinascendingorderbyvariableEPS.Thesamplesfromtwoindustriesafterrandomareidentifiedintable1and2respectively.
Table1:
LeisureIndustryRandomSamples
Table2:
PharmaceuticalIndustryRandomSamples
4.HistogramsandTestofNormality
Manyofthestatisticaltechniques(e.g.parametricstatistics)assumethatthedistributionofscoresonthedependentvariableis“normal”(Pallant,2005).Inotherwords,ifthedistributionofscoresisnotnormal,itneedstochoosedifferentstatisticalteststoanalyseissuessuchascorrelation,differencesbetweenvariables.Asaresult,testofnormalityisoneofthemostimportantstepsinthewholetestsanditcanbeassessedbothbyhistogramandtheExploreoptioninSPSS.
Histograms
AsGravetterandWallnau(2000,pp52)indicated,normalisusedtodescribeasymmetrical,bell-shapednormalcurve,whichhasthegreatestfrequencyofscoresinthemiddle,withsmallerfrequenciestowardstheextremes.
Figure1:
Figure2:
HistogramofEPSin“Leisure”HistogramofGearingin“Leisure”
Figure3:
Figure4:
HistogramofEPSin“Pharmaceutical”HistogramofGearingin“Pharmaceutical”
Figure1and3displaythedistributionsofEPSin“Leisure”and“Pharmaceutical”andtheothertwofiguresexhibitthedistributionofGearingin“Leisure”and“Pharmaceutical”separately.Byobservingthenormalcurveandfrequencyinthose4histograms,itcouldeasilyfindthatthescoresinfourfiguresareallskewedtotheleftandmostscorearenotoccurringinthecentre,taperingouttowardstheextremes,althoughitseemsthatthetwocurvesofEPSandGearingin“Pharmaceutical”alittlebitmorenormalthanthatofthemin“Leisure”.Inordertocertifythattheyareexactlyunnormaldistributed,itneedstotaketestbyExploreoptioninSPSS.
TestofNormality
TheresultsofTestofNormalityaregivenbytheKolmogorov-SmirnovstatisticanditindicatesnormalitywhentheSig.valueismorethan.05(Pallent,2005).AsTable3indicatesthattheSig.values(referto0.002,0.006,0.000,and0.000)arealllessthan.05,thereforethescoresofEPSandGearingeitherin“Leisure”or“Pharmaceutical”arenotnormallydistributed.
Table3:
ThetestofNormalityofEPSandGearingintheTwoSectors
5.CorrelationAnalysis
“Correlationanalysisisusedtodescribethestrengthanddirectionofthelinearrelationshipbetweentwovariables.”(Pallant,2005,pp121)Inthisreport,itneedstoworkoutifthereisarelationshipbetweenEPSandGearingeitherin“Leisure”or“Pharmaceutical”ornot.
Accordingtotheresultoftestofnormalityabove,itdemonstratesthatallthescoresareunnormallydistributed.Therefore,thiscorrelationanalysisneedstotake“Spearmanrankordercorrelation”whichbelongstonon-parametrictechniques.Althoughnon-parametrictechniquearelesspowerfulthanparametricalternative,itisalsousefulwhenthereisrelativesmallamountofsamples(e.g.N=30)andthescoresarenotnormallydistributed.
FormulateHypotheses
In“Leisure”:
Nullhypotheses(H0):
ThereisnocorrelationbetweenEPSandGearingin“Leisure”.
Alternativehypotheses(H1):
ThereisacorrelationbetweenEPSandGearingin“Pharmaceutical”.
IfSig.>.05,H0isacceptedbutH1isrefused;
Sig.<.05,H0isrefusedbutH1isaccepted.
In“Pharmaceutical”:
Nullhypotheses(H0):
ThereisnocorrelationbetweenEPSandGearingin“Pharmaceutical”.
Alternativehypotheses(H1):
ThereisacorrelationbetweenEPSandGearingin“Pharmaceutical”.
IfSig.>.05,H0isacceptedbutH1isrefused;
Sig.<.05,H0isrefusedbutH1isaccepted.
ScatterPlots
BeforedoingacorrelationtestitisagoodideatogenerateascatterplotfirsttoseethenatureoftherelationshipbetweenEPSandGearing(Pallant,2005).
Figure5:
ScatterplotsofEPSandGearingin“Leisure”
Figure6:
ScatterplotsofEPSandGearingin“Pharmaceutical”
ThedatapointsinFigure5andFigure6showsthedifferentsituations.WhilethescatterplotsforoutlierswhichspreadallovertheplaceinFigure5suggestsaverylowcorrelationbetweenEPSandGearing,thepointsinFigure6whichareneatlyarrangedinanarrowcigarshapesuggestquiteastrongcorrelation.Thus,bycheckingthescatterplots,itsuggeststhatthereisasmallrelationshipofEPSandGearingin“Leisure”,nevertheless,thereisasignificantrelationshipofthesetwovariablesin“Pharmaceutical”.
Spearman’sRankOrderCorrelation
Althoughthescatterplotsshow