Study title

One-stage stratified or systematic random sample

The sample required for this inquiry was one of 10,000 civilian adults aged 18 years and over in Great Britain. A two-stage sampling design was used. The first-stage sampling units were local government administrative districts. These were first stratified by Civil Defence Region and two of these Regions were further divided, Wales into two parts (North and South) and Greater London into seven geographical zones. Within Region or Sub-region districts were then stratified by whether urban or rural. The districts in each of the resulting strata were then used in descending order of size (population) and all districts of 140,000 population or over were extracted and each allocated its due proportion of the total sample of 10,000 (districts of this size or over accounted for about 23% of the total sample, a proportion which thus constituted a single-stage stratified random sample of the populations of the larger districts). The remaining districts in each stratum were then grouped into sub-strata by population size and two were selected at random from each sub-stratum. In each of the resulting 225 districts the required sample of civilians aged 18 years and over was obtained by drawing cards at constant intervals from the adult portion of the local National Register.