You are on page 1of 26

8/26/2014 Statistical Tests in SPSS

http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 1/26
givingagift HelptheStatConsultingGroupby
WhatstatisticalanalysisshouldIuse?
StatisticalanalysesusingSPSS
Introduction
ThispageshowshowtoperformanumberofstatisticaltestsusingSPSS.Eachsectiongivesabriefdescriptionoftheaim
ofthestatisticaltest,whenitisused,anexampleshowingtheSPSScommandsandSPSS(oftenabbreviated)outputwitha
briefinterpretationoftheoutput.YoucanseethepageChoosingtheCorrectStatisticalTestforatablethatshowsan
overviewofwheneachtestisappropriatetouse.Indecidingwhichtestisappropriatetouse,itisimportanttoconsiderthe
typeofvariablesthatyouhave(i.e.,whetheryourvariablesarecategorical,ordinalorintervalandwhethertheyarenormally
distributed),seeWhatisthedifferencebetweencategorical,ordinalandintervalvariables?formoreinformationonthis.
Aboutthehsbdatafile
Mostoftheexamplesinthispagewilluseadatafilecalledhsb2,highschoolandbeyond.Thisdatafilecontains200
observationsfromasampleofhighschoolstudentswithdemographicinformationaboutthestudents,suchastheirgender
(female),socioeconomicstatus(ses)andethnicbackground(race).Italsocontainsanumberofscoresonstandardized
tests,includingtestsofreading(read),writing(write),mathematics(math)andsocialstudies(socst).Youcangetthehsb
datafilebyclickingonhsb2.
Onesamplettest
Aonesamplettestallowsustotestwhetherasamplemean(ofanormallydistributedintervalvariable)significantlydiffers
fromahypothesizedvalue.Forexample,usingthehsb2datafile,saywewishtotestwhethertheaveragewritingscore
(write)differssignificantlyfrom50.Wecandothisasshownbelow.
t-test
/testval = 50
/variable = write.
Themeanofthevariablewriteforthisparticularsampleofstudentsis52.775,whichisstatisticallysignificantlydifferentfrom
thetestvalueof50.Wewouldconcludethatthisgroupofstudentshasasignificantlyhighermeanonthewritingtestthan
50.
Onesamplemediantest
Aonesamplemediantestallowsustotestwhetherasamplemediandifferssignificantlyfromahypothesizedvalue.Wewill
usethesamevariable,write,aswedidintheonesamplettestexampleabove,butwedonotneedtoassumethatitis
intervalandnormallydistributed(weonlyneedtoassumethatwriteisanordinalvariable).
nptests
/onesample test (write) wilcoxon(testvalue = 50).
>whatstat.htm > stat > spss whatstat
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 2/26
Binomialtest
Aonesamplebinomialtestallowsustotestwhethertheproportionofsuccessesonatwolevelcategoricaldependent
variablesignificantlydiffersfromahypothesizedvalue.Forexample,usingthehsb2datafile,saywewishtotestwhether
theproportionoffemales(female)differssignificantlyfrom50%,i.e.,from.5.Wecandothisasshownbelow.
npar tests
/binomial (.5) = female.
Theresultsindicatethatthereisnostatisticallysignificantdifference(p=.229).Inotherwords,theproportionoffemalesin
thissampledoesnotsignificantlydifferfromthehypothesizedvalueof50%.
Chisquaregoodnessoffit
Achisquaregoodnessoffittestallowsustotestwhethertheobservedproportionsforacategoricalvariabledifferfrom
hypothesizedproportions.Forexample,let'ssupposethatwebelievethatthegeneralpopulationconsistsof10%Hispanic,
10%Asian,10%AfricanAmericanand70%Whitefolks.Wewanttotestwhethertheobservedproportionsfromoursample
differsignificantlyfromthesehypothesizedproportions.
npar test
/chisquare = race
/expected = 10 10 10 70.
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 3/26
Theseresultsshowthatracialcompositioninoursampledoesnotdiffersignificantlyfromthehypothesizedvaluesthatwe
supplied(chisquarewiththreedegreesoffreedom=5.029,p=.170).
Twoindependentsamplesttest
Anindependentsamplesttestisusedwhenyouwanttocomparethemeansofanormallydistributedintervaldependent
variablefortwoindependentgroups.Forexample,usingthehsb2datafile,saywewishtotestwhetherthemeanforwriteis
thesameformalesandfemales.
t-test groups = female(0 1)
/variables = write.
Becausethestandarddeviationsforthetwogroupsaresimilar(10.3and8.1),wewillusethe"equalvariancesassumed"
test.Theresultsindicatethatthereisastatisticallysignificantdifferencebetweenthemeanwritingscoreformalesand
females(t=3.734,p=.000).Inotherwords,femaleshaveastatisticallysignificantlyhighermeanscoreonwriting(54.99)
thanmales(50.12).
Seealso
SPSSLearningModule:AnoverviewofstatisticaltestsinSPSS
WilcoxonMannWhitneytest
TheWilcoxonMannWhitneytestisanonparametricanalogtotheindependentsamplesttestandcanbeusedwhenyoudo
notassumethatthedependentvariableisanormallydistributedintervalvariable(youonlyassumethatthevariableisat
leastordinal).YouwillnoticethattheSPSSsyntaxfortheWilcoxonMannWhitneytestisalmostidenticaltothatofthe
independentsamplesttest.Wewillusethesamedatafile(thehsb2datafile)andthesamevariablesinthisexampleaswe
didintheindependentttestexampleaboveandwillnotassumethatwrite,ourdependentvariable,isnormallydistributed.
npar test
/m-w = write by female(0 1).
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 4/26
Theresultssuggestthatthereisastatisticallysignificantdifferencebetweentheunderlyingdistributionsofthewritescores
ofmalesandthewritescoresoffemales(z=3.329,p=0.001).
Seealso
FAQ:WhyistheMannWhitneysignificantwhenthemediansareequal?
Chisquaretest
Achisquaretestisusedwhenyouwanttoseeifthereisarelationshipbetweentwocategoricalvariables.InSPSS,the
chisqoptionisusedonthestatisticssubcommandofthecrosstabscommandtoobtaintheteststatisticanditsassociated
pvalue.Usingthehsb2datafile,let'sseeifthereisarelationshipbetweenthetypeofschoolattended(schtyp)and
students'gender(female).Rememberthatthechisquaretestassumesthattheexpectedvalueforeachcellisfiveorhigher.
Thisassumptioniseasilymetintheexamplesbelow.However,ifthisassumptionisnotmetinyourdata,pleaseseethe
sectiononFisher'sexacttestbelow.
crosstabs
/tables = schtyp by female
/statistic = chisq.
Theseresultsindicatethatthereisnostatisticallysignificantrelationshipbetweenthetypeofschoolattendedandgender
(chisquarewithonedegreeoffreedom=0.047,p=0.828).
Let'slookatanotherexample,thistimelookingatthelinearrelationshipbetweengender(female)andsocioeconomicstatus
(ses).Thepointofthisexampleisthatone(orboth)variablesmayhavemorethantwolevels,andthatthevariablesdonot
havetohavethesamenumberoflevels.Inthisexample,femalehastwolevels(maleandfemale)andseshasthreelevels
(low,mediumandhigh).
crosstabs
/tables = female by ses
/statistic = chisq.
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 5/26
Againwefindthatthereisnostatisticallysignificantrelationshipbetweenthevariables(chisquarewithtwodegreesof
freedom=4.577,p=0.101).
Seealso
SPSSLearningModule:AnOverviewofStatisticalTestsinSPSS
Fisher'sexacttest
TheFisher'sexacttestisusedwhenyouwanttoconductachisquaretestbutoneormoreofyourcellshasanexpected
frequencyoffiveorless.Rememberthatthechisquaretestassumesthateachcellhasanexpectedfrequencyoffiveor
more,buttheFisher'sexacttesthasnosuchassumptionandcanbeusedregardlessofhowsmalltheexpectedfrequency
is.InSPSSunlessyouhavetheSPSSExactTestModule,youcanonlyperformaFisher'sexacttestona2x2table,and
theseresultsarepresentedbydefault.Pleaseseetheresultsfromthechisquaredexampleabove.
OnewayANOVA
Aonewayanalysisofvariance(ANOVA)isusedwhenyouhaveacategoricalindependentvariable(withtwoormore
categories)andanormallydistributedintervaldependentvariableandyouwishtotestfordifferencesinthemeansofthe
dependentvariablebrokendownbythelevelsoftheindependentvariable.Forexample,usingthehsb2datafile,saywewish
totestwhetherthemeanofwritediffersbetweenthethreeprogramtypes(prog).Thecommandforthistestwouldbe:
oneway write by prog.
Themeanofthedependentvariabledifferssignificantlyamongthelevelsofprogramtype.However,wedonotknowifthe
differenceisbetweenonlytwoofthelevelsorallthreeofthelevels.(TheFtestfortheModelisthesameastheFtestfor
progbecauseprogwastheonlyvariableenteredintothemodel.Ifothervariableshadalsobeenentered,theFtestforthe
Modelwouldhavebeendifferentfromprog.)Toseethemeanofwriteforeachlevelofprogramtype,
means tables = write by prog.
Fromthiswecanseethatthestudentsintheacademicprogramhavethehighestmeanwritingscore,whilestudentsinthe
vocationalprogramhavethelowest.
Seealso
SPSSTextbookExamples:DesignandAnalysis,Chapter7
SPSSTextbookExamples:AppliedRegressionAnalysis,Chapter8
SPSSFAQ:HowcanIdoANOVAcontrastsinSPSS?
SPSSLibrary:UnderstandingandInterpretingParameterEstimatesinRegressionandANOVA
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 6/26
KruskalWallistest
TheKruskalWallistestisusedwhenyouhaveoneindependentvariablewithtwoormorelevelsandanordinaldependent
variable.Inotherwords,itisthenonparametricversionofANOVAandageneralizedformoftheMannWhitneytestmethod
sinceitpermitstwoormoregroups.WewillusethesamedatafileastheonewayANOVAexampleabove(thehsb2data
file)andthesamevariablesasintheexampleabove,butwewillnotassumethatwriteisanormallydistributedinterval
variable.
npar tests
/k-w = write by prog (1,3).
Ifsomeofthescoresreceivetiedranks,thenacorrectionfactorisused,yieldingaslightlydifferentvalueofchisquared.
Withorwithoutties,theresultsindicatethatthereisastatisticallysignificantdifferenceamongthethreetypeofprograms.
Pairedttest
Apaired(samples)ttestisusedwhenyouhavetworelatedobservations(i.e.,twoobservationspersubject)andyouwantto
seeifthemeansonthesetwonormallydistributedintervalvariablesdifferfromoneanother.Forexample,usingthehsb2
datafilewewilltestwhetherthemeanofreadisequaltothemeanofwrite.
t-test pairs = read with write (paired).
Theseresultsindicatethatthemeanofreadisnotstatisticallysignificantlydifferentfromthemeanofwrite(t=0.867,p=
0.387).
Wilcoxonsignedranksumtest
TheWilcoxonsignedranksumtestisthenonparametricversionofapairedsamplesttest.YouusetheWilcoxonsigned
ranksumtestwhenyoudonotwishtoassumethatthedifferencebetweenthetwovariablesisintervalandnormally
distributed(butyoudoassumethedifferenceisordinal).Wewillusethesameexampleasabove,butwewillnotassumethat
thedifferencebetweenreadandwriteisintervalandnormallydistributed.
npar test
/wilcoxon = write with read (paired).
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 7/26
Theresultssuggestthatthereisnotastatisticallysignificantdifferencebetweenreadandwrite.
Ifyoubelievethedifferencesbetweenreadandwritewerenotordinalbutcouldmerelybeclassifiedaspositiveand
negative,thenyoumaywanttoconsiderasigntestinlieuofsignranktest.Again,wewillusethesamevariablesinthis
exampleandassumethatthisdifferenceisnotordinal.
npar test
/sign = read with write (paired).
Weconcludethatnostatisticallysignificantdifferencewasfound(p=.556).
McNemartest
YouwouldperformMcNemar'stestifyouwereinterestedinthemarginalfrequenciesoftwobinaryoutcomes.Thesebinary
outcomesmaybethesameoutcomevariableonmatchedpairs(likeacasecontrolstudy)ortwooutcomevariablesfroma
singlegroup.Continuingwiththehsb2datasetusedinseveralaboveexamples,letuscreatetwobinaryoutcomesinour
dataset:himathandhiread.Theseoutcomescanbeconsideredinatwowaycontingencytable.Thenullhypothesisisthat
theproportionofstudentsinthehimathgroupisthesameastheproportionofstudentsinhireadgroup(i.e.,thatthe
contingencytableissymmetric).
compute himath = (math>60).
compute hiread = (read>60).
execute.
crosstabs
/tables=himath BY hiread
/statistic=mcnemar
/cells=count.
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 8/26
McNemar'schisquarestatisticsuggeststhatthereisnotastatisticallysignificantdifferenceinthe
proportionofstudentsinthehimathgroupandtheproportionofstudentsinthehireadgroup.
OnewayrepeatedmeasuresANOVA
Youwouldperformaonewayrepeatedmeasuresanalysisofvarianceifyouhadonecategoricalindependentvariableanda
normallydistributedintervaldependentvariablethatwasrepeatedatleasttwiceforeachsubject.Thisistheequivalentofthe
pairedsamplesttest,butallowsfortwoormorelevelsofthecategoricalvariable.Thistestswhetherthemeanofthe
dependentvariablediffersbythecategoricalvariable.Wehaveanexampledatasetcalledrb4wide,whichisusedinKirk's
bookExperimentalDesign.Inthisdataset,yisthedependentvariable,aistherepeatedmeasureandsisthevariablethat
indicatesthesubjectnumber.
glm y1 y2 y3 y4
/wsfactor a(4).
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 9/26
Youwillnoticethatthisoutputgivesfourdifferentpvalues.Theoutputlabeled"sphericityassumed"isthepvalue(0.000)
thatyouwouldgetifyouassumedcompoundsymmetryinthevariancecovariancematrix.Becausethatassumptionisoften
notvalid,thethreeotherpvaluesoffervariouscorrections(theHuynhFeldt,HF,GreenhouseGeisser,GGandLower
bound).Nomatterwhichpvalueyouuse,ourresultsindicatethatwehaveastatisticallysignificanteffectofaatthe.05
level.
Seealso
SPSSTextbookExamplesfromDesignandAnalysis:Chapter16
SPSSLibrary:AdvancedIssuesinUsingandUnderstandingSPSSMANOVA
SPSSCodeFragment:RepeatedMeasuresANOVA
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 10/26
Repeatedmeasureslogisticregression
Ifyouhaveabinaryoutcomemeasuredrepeatedlyforeachsubjectandyouwishtorunalogisticregressionthataccountsfor
theeffectofmultiplemeasuresfromsinglesubjects,youcanperformarepeatedmeasureslogisticregression.InSPSS,this
canbedoneusingtheGENLINcommandandindicatingbinomialastheprobabilitydistributionandlogitasthelinkfunctionto
beusedinthemodel.Theexercisedatafilecontains3pulsemeasurementsfromeachof30peopleassignedto2different
dietregimentsand3differentexerciseregiments.Ifwedefinea"high"pulseasbeingover100,wecanthenpredictthe
probabilityofahighpulseusingdietregiment.
GET FILE='C:\mydata\exercise.sav'.
GENLIN highpulse (REFERENCE=LAST)
BY diet (order = DESCENDING)
/MODEL diet
DISTRIBUTION=BINOMIAL
LINK=LOGIT
/REPEATED SUBJECT=id CORRTYPE = EXCHANGEABLE.
Theseresultsindicatethatdietisnotstatisticallysignificant(WaldChiSquare=1.562,p=0.211).
FactorialANOVA
AfactorialANOVAhastwoormorecategoricalindependentvariables(eitherwithorwithouttheinteractions)andasingle
normallydistributedintervaldependentvariable.Forexample,usingthehsb2datafilewewilllookatwritingscores(write)as
thedependentvariableandgender(female)andsocioeconomicstatus(ses)asindependentvariables,andwewillincludean
interactionoffemalebyses.NotethatinSPSS,youdonotneedtohavetheinteractionterm(s)inyourdataset.Rather,
youcanhaveSPSScreateit/themtemporarilybyplacinganasteriskbetweenthevariablesthatwillmakeuptheinteraction
term(s).
glm write by female ses.
Theseresultsindicatethattheoverallmodelisstatisticallysignificant(F=5.666,p=0.00).Thevariablesfemaleandses
arealsostatisticallysignificant(F=16.595,p=0.000andF=6.611,p=0.002,respectively).However,thatinteraction
betweenfemaleandsesisnotstatisticallysignificant(F=0.133,p=0.875).
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 11/26
Seealso
SPSSTextbookExamplesfromDesignandAnalysis:Chapter10
SPSSFAQ:HowcanIdotestsofsimplemaineffectsinSPSS?
SPSSFAQ:HowdoIplotANOVAcellmeansinSPSS?
SPSSLibrary:AnOverviewofSPSSGLM
Friedmantest
YouperformaFriedmantestwhenyouhaveonewithinsubjectsindependentvariablewithtwoormorelevelsanda
dependentvariablethatisnotintervalandnormallydistributed(butatleastordinal).Wewillusethistesttodetermineifthere
isadifferenceinthereading,writingandmathscores.Thenullhypothesisinthistestisthatthedistributionoftheranksof
eachtypeofscore(i.e.,reading,writingandmath)arethesame.ToconductaFriedmantest,thedataneedtobeinalong
format.SPSShandlesthisforyou,butinotherstatisticalpackagesyouwillhavetoreshapethedatabeforeyoucan
conductthistest.
npar tests
/friedman = read write math.
Friedman'schisquarehasavalueof0.645andapvalueof0.724andisnotstatisticallysignificant.Hence,thereisno
evidencethatthedistributionsofthethreetypesofscoresaredifferent.
Orderedlogisticregression
Orderedlogisticregressionisusedwhenthedependentvariableisordered,butnotcontinuous.Forexample,usingthehsb2
datafilewewillcreateanorderedvariablecalledwrite3.Thisvariablewillhavethevalues1,2and3,indicatingalow,
mediumorhighwritingscore.Wedonotgenerallyrecommendcategorizingacontinuousvariableinthiswaywearesimply
creatingavariabletouseforthisexample.Wewillusegender(female),readingscore(read)andsocialstudiesscore
(socst)aspredictorvariablesinthismodel.Wewillusealogitlinkandontheprintsubcommandwehaverequestedthe
parameterestimates,the(model)summarystatisticsandthetestoftheparallellinesassumption.
if write ge 30 and write le 48 write3 = 1.
if write ge 49 and write le 57 write3 = 2.
if write ge 58 and write le 70 write3 = 3.
execute.
plum write3 with female read socst
/link = logit
/print = parameter summary tparallel.
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 12/26
Theresultsindicatethattheoverallmodelisstatisticallysignificant(p<.000),asareeachofthepredictorvariables(p<
.000).Therearetwothresholdsforthismodelbecausetherearethreelevelsoftheoutcomevariable.Wealsoseethatthe
testoftheproportionaloddsassumptionisnonsignificant(p=.563).Oneoftheassumptionsunderlyingordinallogistic(and
ordinalprobit)regressionisthattherelationshipbetweeneachpairofoutcomegroupsisthesame.Inotherwords,ordinal
logisticregressionassumesthatthecoefficientsthatdescribetherelationshipbetween,say,thelowestversusallhigher
categoriesoftheresponsevariablearethesameasthosethatdescribetherelationshipbetweenthenextlowestcategory
andallhighercategories,etc.Thisiscalledtheproportionaloddsassumptionortheparallelregressionassumption.
Becausetherelationshipbetweenallpairsofgroupsisthesame,thereisonlyonesetofcoefficients(onlyonemodel).If
thiswasnotthecase,wewouldneeddifferentmodels(suchasageneralizedorderedlogitmodel)todescribetherelationship
betweeneachpairofoutcomegroups.
Seealso
SPSSDataAnalysisExamples:Orderedlogisticregression
SPSSAnnotatedOutput:OrdinalLogisticRegression
Factoriallogisticregression
Afactoriallogisticregressionisusedwhenyouhavetwoormorecategoricalindependentvariablesbutadichotomous
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 13/26
dependentvariable.Forexample,usingthehsb2datafilewewillusefemaleasourdependentvariable,becauseitisthe
onlydichotomousvariableinourdatasetcertainlynotbecauseitcommonpracticetousegenderasanoutcomevariable.
Wewillusetypeofprogram(prog)andschooltype(schtyp)asourpredictorvariables.Becauseprogisacategorical
variable(ithasthreelevels),weneedtocreatedummycodesforit.SPSSwilldothisforyoubymakingdummycodesforall
variableslistedafterthekeywordwith.SPSSwillalsocreatetheinteractiontermsimplylistthetwovariablesthatwillmake
uptheinteractionseparatedbythekeywordby.
logistic regression female with prog schtyp prog by schtyp
/contrast(prog) = indicator(1).
Theresultsindicatethattheoverallmodelisnotstatisticallysignificant(LRchi2=3.147,p=0.677).Furthermore,noneof
thecoefficientsarestatisticallysignificanteither.Thisshowsthattheoveralleffectofprogisnotsignificant.
Seealso
Annotatedoutputforlogisticregression
SPSSTopics:LogisticRegression
Correlation
Acorrelationisusefulwhenyouwanttoseetherelationshipbetweentwo(ormore)normallydistributedintervalvariables.For
example,usingthehsb2datafilewecanrunacorrelationbetweentwocontinuousvariables,readandwrite.
correlations
/variables = read write.
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 14/26
Inthesecondexample,wewillrunacorrelationbetweenadichotomousvariable,female,andacontinuousvariable,write.
Althoughitisassumedthatthevariablesareintervalandnormallydistributed,wecanincludedummyvariableswhen
performingcorrelations.
correlations
/variables = female write.
Inthefirstexampleabove,weseethatthecorrelationbetweenreadandwriteis0.597.Bysquaringthecorrelationandthen
multiplyingby100,youcandeterminewhatpercentageofthevariabilityisshared.Let'sround0.597tobe0.6,whichwhen
squaredwouldbe.36,multipliedby100wouldbe36%.Hencereadsharesabout36%ofitsvariabilitywithwrite.Inthe
outputforthesecondexample,wecanseethecorrelationbetweenwriteandfemaleis0.256.Squaringthisnumberyields
.065536,meaningthatfemalesharesapproximately6.5%ofitsvariabilitywithwrite.
Seealso
Annotatedoutputforcorrelation
SPSSLearningModule:AnOverviewofStatisticalTestsinSPSS
SPSSFAQ:HowcanIanalyzemydatabycategories?
MissingDatainSPSS
Simplelinearregression
Simplelinearregressionallowsustolookatthelinearrelationshipbetweenonenormallydistributedintervalpredictorandone
normallydistributedintervaloutcomevariable.Forexample,usingthehsb2datafile,saywewishtolookattherelationship
betweenwritingscores(write)andreadingscores(read)inotherwords,predictingwritefromread.
regression variables = write read
/dependent = write
/method = enter.
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 15/26
Weseethattherelationshipbetweenwriteandreadispositive(.552)andbasedonthetvalue(10.47)andpvalue(0.000),
wewouldconcludethisrelationshipisstatisticallysignificant.Hence,wewouldsaythereisastatisticallysignificantpositive
linearrelationshipbetweenreadingandwriting.
Seealso
RegressionWithSPSS:Chapter1SimpleandMultipleRegression
Annotatedoutputforregression
SPSSTopics:Regression
SPSSTextbookExamples:IntroductiontothePracticeofStatistics,Chapter10
SPSSTextbookExamples:RegressionwithGraphics,Chapter2
SPSSTextbookExamples:AppliedRegressionAnalysis,Chapter5
Nonparametriccorrelation
ASpearmancorrelationisusedwhenoneorbothofthevariablesarenotassumedtobenormallydistributedandinterval(but
areassumedtobeordinal).Thevaluesofthevariablesareconvertedinranksandthencorrelated.Inourexample,wewill
lookforarelationshipbetweenreadandwrite.Wewillnotassumethatbothofthesevariablesarenormalandinterval.
nonpar corr
/variables = read write
/print = spearman.
Theresultssuggestthattherelationshipbetweenreadandwrite(rho=0.617,p=0.000)isstatisticallysignificant.
Simplelogisticregression
Logisticregressionassumesthattheoutcomevariableisbinary(i.e.,codedas0and1).Wehaveonlyonevariableinthe
hsb2datafilethatiscoded0and1,andthatisfemale.Weunderstandthatfemaleisasillyoutcomevariable(itwould
makemoresensetouseitasapredictorvariable),butwecanusefemaleastheoutcomevariabletoillustratehowthecode
forthiscommandisstructuredandhowtointerprettheoutput.Thefirstvariablelistedafterthelogisticcommandisthe
outcome(ordependent)variable,andalloftherestofthevariablesarepredictor(orindependent)variables.Inourexample,
femalewillbetheoutcomevariable,andreadwillbethepredictorvariable.AswithOLSregression,thepredictorvariables
mustbeeitherdichotomousorcontinuoustheycannotbecategorical.
logistic regression female with read.
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 16/26
Theresultsindicatethatreadingscore(read)isnotastatisticallysignificantpredictorofgender(i.e.,beingfemale),Wald=
.562,p=0.453.Likewise,thetestoftheoverallmodelisnotstatisticallysignificant,LRchisquared0.56,p=0.453.
Seealso
Annotatedoutputforlogisticregression
SPSSTopics:LogisticRegression
SPSSLibrary:Whatkindofcontrastsarethese?
Multipleregression
Multipleregressionisverysimilartosimpleregression,exceptthatinmultipleregressionyouhavemorethanonepredictor
variableintheequation.Forexample,usingthehsb2datafilewewillpredictwritingscorefromgender(female),reading,
math,scienceandsocialstudies(socst)scores.
regression variable = write female read math science socst
/dependent = write
/method = enter.
Theresultsindicatethattheoverallmodelisstatisticallysignificant(F=58.60,p=0.000).Furthermore,allofthepredictor
variablesarestatisticallysignificantexceptforread.
Seealso
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 17/26
RegressionwithSPSS:Chapter1SimpleandMultipleRegression
Annotatedoutputforregression
SPSSTopics:Regression
SPSSFrequentlyAskedQuestions
SPSSTextbookExamples:RegressionwithGraphics,Chapter3
SPSSTextbookExamples:AppliedRegressionAnalysis
Analysisofcovariance
AnalysisofcovarianceislikeANOVA,exceptinadditiontothecategoricalpredictorsyoualsohavecontinuouspredictorsas
well.Forexample,theonewayANOVAexampleusedwriteasthedependentvariableandprogastheindependent
variable.Let'saddreadasacontinuousvariabletothismodel,asshownbelow.
glm write with read by prog.
Theresultsindicatethatevenafteradjustingforreadingscore(read),writingscoresstillsignificantlydifferbyprogramtype
(prog),F=5.867,p=0.003.
Seealso
SPSSTextbookExamplesfromDesignandAnalysis:Chapter14
SPSSLibrary:AnOverviewofSPSSGLM
SPSSLibrary:HowdoIhandleinteractionsofcontinuousandcategoricalvariables?
Multiplelogisticregression
Multiplelogisticregressionislikesimplelogisticregression,exceptthattherearetwoormorepredictors.Thepredictorscan
beintervalvariablesordummyvariables,butcannotbecategoricalvariables.Ifyouhavecategoricalpredictors,theyshould
becodedintooneormoredummyvariables.Wehaveonlyonevariableinourdatasetthatiscoded0and1,andthatis
female.Weunderstandthatfemaleisasillyoutcomevariable(itwouldmakemoresensetouseitasapredictorvariable),
butwecanusefemaleastheoutcomevariabletoillustratehowthecodeforthiscommandisstructuredandhowtointerpret
theoutput.Thefirstvariablelistedafterthelogisticregressioncommandistheoutcome(ordependent)variable,andallof
therestofthevariablesarepredictor(orindependent)variables(listedafterthekeywordwith).Inourexample,femalewillbe
theoutcomevariable,andreadandwritewillbethepredictorvariables.
logistic regression female with read write.
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 18/26
Theseresultsshowthatbothreadandwritearesignificantpredictorsoffemale.
Seealso
Annotatedoutputforlogisticregression
SPSSTopics:LogisticRegression
SPSSTextbookExamples:AppliedLogisticRegression,Chapter2
SPSSCodeFragments:GraphingResultsinLogisticRegression
Discriminantanalysis
Discriminantanalysisisusedwhenyouhaveoneormorenormallydistributedintervalindependentvariablesandacategorical
dependentvariable.Itisamultivariatetechniquethatconsidersthelatentdimensionsintheindependentvariablesfor
predictinggroupmembershipinthecategoricaldependentvariable.Forexample,usingthehsb2datafile,saywewishtouse
read,writeandmathscorestopredictthetypeofprogramastudentbelongsto(prog).
discriminate groups = prog(1, 3)
/variables = read write math.
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 19/26
Clearly,theSPSSoutputforthisprocedureisquitelengthy,anditisbeyondthescopeofthispagetoexplainallofit.
However,themainpointisthattwocanonicalvariablesareidentifiedbytheanalysis,thefirstofwhichseemstobemore
relatedtoprogramtypethanthesecond.
Seealso
discriminantfunctionanalysis
SPSSLibrary:AHistoryofSPSSStatisticalFeatures
OnewayMANOVA
MANOVA(multivariateanalysisofvariance)islikeANOVA,exceptthattherearetwoormoredependentvariables.Inaone
wayMANOVA,thereisonecategoricalindependentvariableandtwoormoredependentvariables.Forexample,usingthe
hsb2datafile,saywewishtoexaminethedifferencesinread,writeandmathbrokendownbyprogramtype(prog).
glm read write math by prog.
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 20/26
Thestudentsinthedifferentprogramsdifferintheirjointdistributionofread,writeandmath.
Seealso
SPSSLibrary:AdvancedIssuesinUsingandUnderstandingSPSSMANOVA
GLM:MANOVAandMANCOVA
SPSSLibrary:MANOVAandGLM
Multivariatemultipleregression
Multivariatemultipleregressionisusedwhenyouhavetwoormoredependentvariablesthataretobepredictedfromtwoor
moreindependentvariables.Inourexample,wewillpredictwriteandreadfromfemale,math,scienceandsocialstudies
(socst)scores.
glm write read with female math science socst.
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 21/26
Theseresultsshowthatallofthevariablesinthemodelhaveastatisticallysignificantrelationshipwiththejointdistribution
ofwriteandread.
Canonicalcorrelation
Canonicalcorrelationisamultivariatetechniqueusedtoexaminetherelationshipbetweentwogroupsofvariables.Foreach
setofvariables,itcreateslatentvariablesandlooksattherelationshipsamongthelatentvariables.Itassumesthatall
variablesinthemodelareintervalandnormallydistributed.SPSSrequiresthateachofthetwogroupsofvariablesbe
separatedbythekeywordwith.Thereneednotbeanequalnumberofvariablesinthetwogroups(beforeandafterthewith).
manova read write with math science
/discrim.
* * * * * * A n a l y s i s o f V a r i a n c e -- design 1 * * * * * *
EFFECT .. WITHIN CELLS Regression
Multivariate Tests of Significance (S = 2, M = -1/2, N = 97 )
Test Name Value Approx. F Hypoth. DF Error DF Sig. of F
Pillais .59783 41.99694 4.00 394.00 .000
Hotellings 1.48369 72.32964 4.00 390.00 .000
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 22/26
Wilks .40249 56.47060 4.00 392.00 .000
Roys .59728
Note.. F statistic for WILKS' Lambda is exact.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
EFFECT .. WITHIN CELLS Regression (Cont.)
Univariate F-tests with (2,197) D. F.
Variable Sq. Mul. R Adj. R-sq. Hypoth. MS Error MS F
READ .51356 .50862 5371.66966 51.65523 103.99081
WRITE .43565 .42992 3894.42594 51.21839 76.03569
Variable Sig. of F
READ .000
WRITE .000
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Raw canonical coefficients for DEPENDENT variables
Function No.
Variable 1
READ .063
WRITE .049
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Standardized canonical coefficients for DEPENDENT variables
Function No.
Variable 1
READ .649
WRITE .467
* * * * * * A n a l y s i s o f V a r i a n c e -- design 1 * * * * * *
Correlations between DEPENDENT and canonical variables
Function No.
Variable 1
READ .927
WRITE .854
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Variance in dependent variables explained by canonical variables
CAN. VAR. Pct Var DE Cum Pct DE Pct Var CO Cum Pct CO
1 79.441 79.441 47.449 47.449
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Raw canonical coefficients for COVARIATES
Function No.
COVARIATE 1
MATH .067
SCIENCE .048
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Standardized canonical coefficients for COVARIATES
CAN. VAR.
COVARIATE 1
MATH .628
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 23/26
SCIENCE .478
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Correlations between COVARIATES and canonical variables
CAN. VAR.
Covariate 1
MATH .929
SCIENCE .873
* * * * * * A n a l y s i s o f V a r i a n c e -- design 1 * * * * * *
Variance in covariates explained by canonical variables
CAN. VAR. Pct Var DE Cum Pct DE Pct Var CO Cum Pct CO
1 48.544 48.544 81.275 81.275
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Regression analysis for WITHIN CELLS error term
--- Individual Univariate .9500 confidence intervals
Dependent variable .. READ reading score
COVARIATE B Beta Std. Err. t-Value Sig. of t
MATH .48129 .43977 .070 6.868 .000
SCIENCE .36532 .35278 .066 5.509 .000
COVARIATE Lower -95% CL- Upper
MATH .343 .619
SCIENCE .235 .496
Dependent variable .. WRITE writing score
COVARIATE B Beta Std. Err. t-Value Sig. of t
MATH .43290 .42787 .070 6.203 .000
SCIENCE .28775 .30057 .066 4.358 .000
COVARIATE Lower -95% CL- Upper
MATH .295 .571
SCIENCE .158 .418
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
* * * * * * A n a l y s i s o f V a r i a n c e -- design 1 * * * * * *
EFFECT .. CONSTANT
Multivariate Tests of Significance (S = 1, M = 0, N = 97 )
Test Name Value Exact F Hypoth. DF Error DF Sig. of F
Pillais .11544 12.78959 2.00 196.00 .000
Hotellings .13051 12.78959 2.00 196.00 .000
Wilks .88456 12.78959 2.00 196.00 .000
Roys .11544
Note.. F statistics are exact.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
EFFECT .. CONSTANT (Cont.)
Univariate F-tests with (1,197) D. F.
Variable Hypoth. SS Error SS Hypoth. MS Error MS F Sig. of F
READ 336.96220 10176.0807 336.96220 51.65523 6.52329 .011
WRITE 1209.88188 10090.0231 1209.88188 51.21839 23.62202 .000
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 24/26
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
EFFECT .. CONSTANT (Cont.)
Raw discriminant function coefficients
Function No.
Variable 1
READ .041
WRITE .124
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Standardized discriminant function coefficients
Function No.
Variable 1
READ .293
WRITE .889
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Estimates of effects for canonical variables
Canonical Variable
Parameter 1
1 2.196
* * * * * * A n a l y s i s o f V a r i a n c e -- design 1 * * * * * *
EFFECT .. CONSTANT (Cont.)
Correlations between DEPENDENT and canonical variables
Canonical Variable
Variable 1
READ .504
WRITE .959
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Theoutputaboveshowsthelinearcombinationscorrespondingtothefirstcanonicalcorrelation.Atthebottomoftheoutput
arethetwocanonicalcorrelations.Theseresultsindicatethatthefirstcanonicalcorrelationis.7728.TheFtestinthis
outputteststhehypothesisthatthefirstcanonicalcorrelationisequaltozero.Clearly,F=56.4706isstatistically
significant.However,thesecondcanonicalcorrelationof.0235isnotstatisticallysignificantlydifferentfromzero(F=
0.1087,p=0.7420).
Factoranalysis
Factoranalysisisaformofexploratorymultivariateanalysisthatisusedtoeitherreducethenumberofvariablesinamodel
ortodetectrelationshipsamongvariables.Allvariablesinvolvedinthefactoranalysisneedtobeintervalandareassumed
tobenormallydistributed.Thegoaloftheanalysisistotrytoidentifyfactorswhichunderliethevariables.Theremaybe
fewerfactorsthanvariables,buttheremaynotbemorefactorsthanvariables.Forourexample,let'ssupposethatwethink
thattherearesomecommonfactorsunderlyingthevarioustestscores.Wewillincludesubcommandsforvarimaxrotation
andaplotoftheeigenvalues.Wewilluseaprincipalcomponentsextractionandwillretaintwofactors.(Usingtheseoptions
willmakeourresultscompatiblewiththosefromSASandStataandarenotnecessarilytheoptionsthatyouwillwanttouse.)
factor
/variables read write math science socst
/criteria factors(2)
/extraction pc
/rotation varimax
/plot eigen.
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 25/26
Communality(whichistheoppositeofuniqueness)istheproportionofvarianceofthevariable(i.e.,read)thatisaccounted
forbyallofthefactorstakentogether,andaverylowcommunalitycanindicatethatavariablemaynotbelongwithanyofthe
factors.Thescreeplotmaybeusefulindetermininghowmanyfactorstoretain.Fromthecomponentmatrixtable,wecan
seethatallfiveofthetestscoresloadontothefirstfactor,whileallfivetendtoloadnotsoheavilyonthesecondfactor.The
purposeofrotatingthefactorsistogetthevariablestoloadeitherveryhighorverylowoneachfactor.Inthisexample,
becauseallofthevariablesloadedontofactor1andnotonfactor2,therotationdidnotaidintheinterpretation.Instead,it
madetheresultsevenmoredifficulttointerpret.
Seealso
SPSSFAQ:WhatdoesCronbach'salphamean?
8/26/2014 Statistical Tests in SPSS
http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm 26/26
Howtocitethispage Reportanerroronthispageorleaveacomment
Thecontentofthiswebsiteshouldnotbeconstruedasanendorsementofanyparticularwebsite,book,orsoftwareproductbytheUniversityof
California.

I D R E R E S E A R C H T E C H N O L O G Y
G R O U P
Hig h Perf ormance Comput ing
St at ist ical Comput ing
GIS and Visualiz at ion
High Performance Computing GIS Statistical Computing
Hoffman2 Cluster Mapshare Classes
Hoffman2 Account Application Visualiz ation Conferences
Hoffman2 Usage Statistics 3D Modeling Reading Materials
UC Grid Portal Technology Sandbox IDRE Listserv
UCLA Grid Portal Tech Sandbox Access IDRE Resources
Shared Cluster & Storage Data Centers Social Sciences Data Archive
About IDRE
ABOUT CONTACT NEWS EVENTS OUREXPERTS

2014 UCRegents TermsofUse &PrivacyPolicy

You might also like