Abstract
For some time transportation analysts have recognised the potential of survey designs which permit the analysis of recurrent travel choice and the assessment of change in travel behaviour over time. The opportunity to conduct such surveys is now upon us with recent longitudinal transport surveys in different countries. This paper has its main objective discussion of a series of issues that precede the analysis of longitudinal data and which are critical to the success of any models estimated on such data. The emphasis is on surveys which involve repeated measurements on the sample of units at different points in times designs which are known as panel designs.