Health Education Research, Vol. 3, No. 3, 257-266, 1988
© 1988 Oxford University Press
research-article |
The presentation of an edited version of smoking behaviour
Social Research Unit, Department of Epidemiology and Social Oncology, Christie Hospital & Holt Radium Institute Kinnaird Road, Withington, Manchester M20 9QL, UK
A longitudinal study of smoking among nurses and student teachers which utilized both questionnaires and interviews provided insight into the problem of measuring smoking behaviour. Some participants sought to present their own definition of the situation by editing their smoking behaviour. Editing included refusal to participate, discrepancies in their accounts of their behaviour and issues relating to differences between there searcher's definition of smoking and that of participants. Editing might also involve redefining behaviour by giving contradictory information at different points in time, and evasiveness. The study indicated how information may be revealed, altered or reinterpreted retrospectively. It also suggested some limitations of a longitudinal design in terms of the expectations of participants and the snap-shot examination of a dynamic behaviour, which was associated with gaps in the data collection. The outcome tended towards underestimation of both prevalence and consumption.