BRITISH MEDICAL JOURNAL VOLUME 287 15 OCTOBER 1983
signed to select about 600 adults per town. In years 2 and 3 sample
sizes were increased to about 1000 adults per town.
Follow up samples (years 2 and 3) consisted of subsamples of those
previously screened plus a new randomly selected sample. This report
considers responses only in relation to smoking for the independent
samples. Further analysis is to be carried out on the responses of people
sampled more than once.
The questionnaire included attitudes to smoking (six questions),
knowledge of the effects of smoking (six questions), and smoking
behaviour (four questions). Interviews were conducted by trained
interviewers, most of whom travelled to each town at each measure-INTERVENTION TECHNIQUES