ABSTRACT
This chapter begins with a reading of the sociodemographic-bias literature and describes an attempt to resolve some contradictions by analysing new micro data. Important too is a distinction between initial, sometimes termed 'interim', and final, or 'terminal', response. The great sensitivity of survey-response behaviour to researcher input means that first results of fieldwork often become altered as input is increased. B. Redpath, when tallying response behaviour against British census records for all sample listings on a Family Expenditure Survey, found high response among married women. Desirable as nationally representative data are, municipal level micro data bring unique advantages to the study of nonresponse bias. Cities maintain public socio-demographic information unobtainable in a national sampling plan. Socio-demographic details on sample listings were obtained from the city directories for Waterloo, Kitchener and Cambridge and from municipal assessment records for each city. The synthesis between input and sociodemographic explanations was achieved by entering total response as an independent variable alongside sociodemographic attributes.
