Great Britain Sample Design

British Social Attitudes survey sample design and timeline

Three versions of the British Social Attitudes questionnaire are usually fielded. Each ‘module’ of questions is asked either of the full sample (around 3,450 respondents) or of a random two-thirds or one-third of the sample. The Mental Health Stigma module will be asked of one-third of the sample (around 1,150 people).

Sample design

The British Social Attitudes survey is designed to yield a representative sample of adults aged 18 or over. Since 1993, the sampling frame for the survey has been the Postcode Address File (PAF), a list of addresses (or postal delivery points) compiled by the Post Office.

For practical reasons, the sample is confined to those living in private households. People living in institutions (though not in private households at such institutions) are excluded, as are households whose addresses were not on the PAF.

The sampling method involved a multi-stage design, with three separate stages of selection.

Selection of sectors

At the first stage, postcode sectors are selected systematically from a list of all postal sectors in Great Britain . Before selection, any sectors with fewer than 500 addresses are identified and grouped together with an adjacent sector; in Scotland all sectors north of the Caledonian Canal are excluded (because of the prohibitive costs of interviewing there). Sectors were then stratified on the basis of:

·       37 sub-regions

·       population density with variable banding used, in order to create three equal-sized strata per sub-region

·       ranking by percentage of homes that were owner-occupied in England and Wales and percentage of homes where the head of household was non-manual in Scotland .

Two hundred postcode sectors are selected, with probability proportional to the number of addresses in each sector.

Selection of addresses

Thirty-one addresses are selected in each of the 200 sectors. The sample is therefore 200 x 31 = 6,200 addresses, selected by starting from a random point on the list of addresses for each sector, and choosing each address at a fixed interval. The fixed interval is calculated for each sector in order to generate the correct number of addresses.

The Multiple-Output Indicator (MOI) available through PAF is used when selecting addresses in Scotland . The MOI shows the number of accommodation spaces sharing one address. Thus, if the MOI indicates more than one accommodation space at a given address, the chances of the given address being selected from the list of addresses is increased so that it matched the total number of accommodation spaces. The MOI is largely irrelevant in England and Wales as separate dwelling units generally appear as separate entries on PAF. In Scotland , tenements with many flats tend to appear as one entry on PAF. However, even in Scotland , the vast majority of MOIs have a value of one. The remainder are incorporated into the weighting procedures (described below).

Selection of individuals

Interviewers call at each address selected from PAF and list all those eligible for inclusion in the sample – that is, all persons currently aged 18 or over and resident at the selected address. The interviewer the selects one respondent using a computer-generated random selection procedure. Where there are two or more households or ‘dwelling units’ at a selected address, interviewers first have to select one household or dwelling unit using the same random procedure. They then follow the same procedure to select a person for interview.

Weighting

Data are weighted to take account of the fact that not all the units covered in the survey have the same probability of selection. The weighting reflects the relative selection probabilities of the individual at the three main stages of selection: address, household and individual.

First, because addresses in Scotland are selected using the MOI, weights have to be applied to compensate for the greater probability of an address with an MOI of more than one being selected, compared to an address with an MOI of one. (This stage is omitted for the English and Welsh data.)

Secondly, data are weighted to compensate for the fact that dwelling units at an address which contain a large number of dwelling units are less likely to be selected for inclusion in the survey than ones which do not share an address. (We use this procedure because in most cases of MOIs greater than one, the two stages will cancel each other out, resulting in more efficient weights.)

Thirdly, data are weighted to compensate for the lower selection probabilities of adults living in large households compared with those living in small households. By way of illustration, in 2002 the weights were capped at 8.0 (causing three cases to have their weights reduced). The mean weight was 1.82. The weights are scaled down to make the number of weighted productive cases exactly equal to the number of unweighted productive cases.

2005 survey timeline

Dec – Mar: questionnaire development and 2 small scale paper pilots of new questions

April - May: questionnaire programming and document preparation

June: personal briefings of interviewers

June - September: main fieldwork

September - November: Cleaning and editing of dataset

December: clean data available

1022 E. Third Street, Bloomington, IN 47405 (812) 855-3841

Last updated: 6 September 2005
Comments: sgcmhs@indiana.edu
Copyright 2004, The Trustees of Indiana University
Copyright Complaints