ABSTRACT
Whiletheabovegroupingisausefulclassificationforourdiscussion,itshouldnotbethoughttoimplythatthemethodscannot becombined.Indeedacombinationofmethodsisoftenappropriate tomakeuseoftheirdifferentstrengths.Gray(1957)describeshow amailquestionnairesurveywassupplementedbyinterviewingto increasetheresponserate.Aspartofanothermailsurvey,Grayand Parr(1959)collectedasampleofcigarettestubsforobservational measurement.Ininterviewsurveys,directobservationalmethodsare oftenvaluable,fortheinterviewercanobtaintheinformationfor herself,ratherthanrelyonthepossiblyinaccuraterepliesgivenby respondents.Thus,forexample,inasurveyofthepostalservices,to findouttheeffectsofmovingletter-boxesfromthefrontdoorsof housestoaboxatthegardengate,interviewersmeasuredthe dimensionsofthepresentboxandpacedoutthedistancefromitto thegate,aswellasinterviewingpeoplefortheirreactionstosucha change(Scott,1959);inasurveyofmotorcycleridersontheuseof crashhelmets,theinterviewernotedwhetherthedriverandpassenger werewearinghelmetswhenstoppedforinterview(Scott,1961b);and inasurveyoftheblindbyGrayandTodd(1967).informationabout opinions,attitudesandmobilityoftheblindwasobtainedby straightforwardinterviewing,withtheinterviewersalsomeasuring readingperformanceatBrailleandMoonbytimingthereadingofa shortpassageandaskingafewquestionstoestablishthatthe passagehadbeenunderstood.
