Testing Nine Hypotheses about Quota Sampling

Catherine Marsh 
University of Manchester 
Elinor Scarbrough
University of Essex


Since Moser and Stuart's pioneering work (Moser and Stuart 1953), systematic investigation of random and quota sampling has fallen into abeyance amongst British researchers.1 In the USA, the development of 'probability sampling with quotas', or PSQ (Sudman 1966, 1976; Stephenson 1978) appears to have put an end to the debate about probability and quota sampling which followed the 1948 Presidential election. In Britain, however, the research community is more divided about sampling practice, with market research agencies largely wedded to quota sampling while academic and government practitioners hold to random sampling.2 What is missing in British survey research is a coherent programme reviewing the effects of different sampling procedures on the character of the resulting samples.