Longitudinal designs typically need to be appropriate to particular policy contexts. This means that the designer will often be faced with difficult prior judgements as to the nature of policy developments and their consequences; together with their likely timescale, and possibly, their geographical diffusion. The current use of longitudinal studies in transport has focused largely upon addressing relatively short-term questions, for example monitoring the responses of individuals to interim policy changes. In the medium to long term, however, if such approaches are to continue to increase our understanding of travel behaviour, other and more fundamental influences are likely to become increasingly important. These relate at one level to movements of individuals or households between sub-populations (whether defined in terms of geographical areas, age groups or household characteristics), and at a higher level to changes in basic demographic variables, such as birth rates, mortality, family formation, unemployment etc. This paper attempts to illustrate and clarify the form which such problems can take, in practice, and makes reference to policy questions which have recently been explored, using longitudinal designs, at transport studies unit, oxford university.(a) for the covering abstract of the conference see IRRD 290118.
Abstract