66 CHAPTER 3 MARKETINGRESEARCH:ANAIDTODECISIONMAKING
- Fsychodrama.Inthepsychodramatictypeofquestion, therespondentis asked to
projecthimself intoanartificialmarketingsituation. Theobviousartificialityof
thesituationmakesthepsychodramaa "role-playing"experimentin whichthe
re:.;pondentprovidesinformationbasedonhisp ersonalattitudesthroughhisexpla-
nationoftheartificialsituation.
Perhapsthegreatestdeficiencyofprojectivetechniquesisthedifficultyofpresent-
ingthefindings.Theidentificationofattitudes,motives,opinions,and:;0forthisnotdif-
ficlllt;however,it isextremelyhardtomeasuretheimportanceofthesefactor~.
SelectingtheSample Inmostmarketingresearch, it is~eldojl1necessarytocon-
ducta completecensus;i .e.,totalkto100%ofthetargetsegment.Todosoi~;time-con-
smningandexpensive.Forthis reasonmostmarketingst:rveysmakeuseofsmaples.Asample
isa groupofelements(persons,stores, financialreports)chosenfromamonga "totalpop-
ulation" or" universe."Thevalueof aresearchprojectisdirectlyaffectedbyhowwellthe
samplehasbeenconceivedandconstructed.s
Theselectionofthesampletobeinvestigatedrequiresa masterlist,ora framework,
fromwhichtheymaybeselected.Thesamplingframeis the"population"orstatistical"uni-
verse"fromwhichthesampleunitswillbeselected.Theframefora surveyof attitudesof
creditcustomersofa departmentstorewouldbethecompany'slistofcustomersusingcharge
accounts.
Althoughtherearemanykindsofsampledesigns,allofthemcanbeclassifiedas
eitherprobabilitysamplesornonprobabilitysamples.Ina probabilitysa:nple,eachunithas
a knownchanceofbeingselectedforinclusioninthesample.~tssimplestversionis the
simplerandomsample, inwhicheachunitinthesampleframehasexactly' hesamechance
ofselection.Examplesofthisii1cludeflippinga faircoin,whosesideshavea 50%chance
ofturningup,andthrowinganunloadeddie,whosesideshavea162/3%chanceofturn-
ingup.Thissameplinciplecanbeappliedtothepreviousdepar:ment::.loreexample.A
sampleofnamescouldbeselectedfromthecompany'slistofchargecustomersaccording
toa randomprocess,suchasthatofusinga tableofrandomdigits.
'WhileinClprobabilitysamplethesamplingunitshavea knownchanceof beingselected,
inanonprobabilitysampiethesamplingunitsareselectedarbitrarily.Toreturntoourdepart-
mentstoreexample,insteadofusinga tableofrandomnumberstoselecta sampleofcharge
customers,anarbitraryandmoreconvenientmethodwouldoetotakethefirstfiftyorsixty
namesonthelist.
AnticipatingtheResultslMaking th eReport Theresearchplanshouldprovide
for:(1)proceduresforprocessingthedata; (2)proceduresforinterpretationandanalysis
ofthefindings;and(3)anoutlineofthefinalreport.Inreachingthesedecisions,itisusu-
allyhelpfultoworkfromtheformandcontentofthefinalreport.Thereportshouldpres-
enta summaryoffindingsandrecommendationsformanagementactiondrawnupinthe
lightofthereasonsfortheresearch.'Thekind~offactstobepresentedandthemannerof
theirpresentationdictatesthetypeofanalysistobeundertaken.Thekindsofanalysiswill ,
inturn,oftensuggestthemethodofdataprocessing.Dataprocessingingeneralrefers to
theproceduresforsorting,assembling,andreportingdata.Itcanbedonemanuallybythe
useofwork sheetsorbycomputerprogramming.Themethodofdataprocessinghasimpor-
tant bearinguponL'1emannerinwhich~hedataarecollectedandreported.Thus, thedesign
oftheprojectis oftenexpeditedbya thoroughconsiderationofthekindsofresultsthatare
expectedandhowtheywillbehandledinthefina:report.
Anticipatingtheresultsofthe~)fojectandpreparinga "dummy"finalreporthasanother
advantage.It is oftenhelpful~ou~etheresultsofthisstepintheresearchdesigntodemon-
