Collecting retrospective data: accuracy of recall after 50 years judged against historical records.

Recent interest in a lifecourse perspective on health inequalities will rekindle concerns about the accuracy of retrospective data. The present paper demonstrates that recalled information on some types of social circumstances can be obtained with a useful degree of accuracy using an interview technique which helps to minimize recall bias. Lifegrid information about social circumstances during their youth and childhood was collected from 57 subjects in early old age and compared with archive material of the same subjects' social circumstances recorded 50 years previously. A comparison of interview with archive data revealed that a substantial majority of subjects had recalled simple socio-demographic information after a period of 50 years with a useful degree of accuracy. Within lifecourse research, it is concluded, carefully collected retrospective data offer a valuable complement to birth cohort studies, provided that such usage is sensitive to the types of items of information which can, and can not, be recalled accurately.