ABSTRACT
I NTHElasttwochapterstheprincipalwaysofcollectingsurveydata,mailquestionnairesandpersonalinterviews,havebeen discussed.Wenowturntotheinstrumentonwhichbothap-
proachesdepend,thequestionnaireorrecordingschedule. Asbrieflynotedabove,thesetwotermsaresometimesusedto
distinguishbetweenthesituationinwhichtherespondenthimself fillsintheanswersandthatinwhichaninterviewerasksthequestions andrecordstheanswers. 1Tothesurveydesignerthedistinctionis important,forinonecaseheisproducingadocumenttobeused and,hehopes,understoodbytherespondentunaided,whileinthe othertheuserswillbepersonsespeciallytrainedtohandlesuch documents.ThequestionnaireusedintheGovernmentSocial Survey'sradioandtelevisionenquiry,whichwasreferredtoin ChapterII,isagoodexamplecftheformertypeandisreproduced onpp.305-7. 2Inthisenquiryanexperimentwascarriedoutonthe wordingofthecoveringletter;halfthesampleweresentarelatively impersonalletterandtheotherhalfarelativelypersonalone.There wasnosignificantdifferenceintheresponserateforthetwohalves ofthesample.Inbothcasestheletterisbrief,butcoverstheessential points:theletterheadingindicatesthesponsorshipoftheenquiry, andthecontentsinformtherespondentsofthesubjectofthe enquiryandofthefactthattheyrepresentacross-sectionofthe populationandtellthemclearlywhattheyareaskedtodo.The questionnaireitselfiscarefullyandclearlylaidout,withboldtype
304 SURVEY METHODS IN SOCIAL INVESTIGATION
and capitals being widely used to emphasize particular words and instructions.1 What is more, the questions themselves are simple and it looks as if a genuine effort was made to keep them to a minimum.