Abstract
Daily mobility is a multidimensional construct. Location sensing enables measuring an individual’s daily mobility in various ways and this has prompted the issue of choosing appropriate mobility indicators for a given application, in particular in the health sciences, where the aim is to link mobility behavior to outcomes related to health and well-being. We previously proposed a classification framework for daily mobility indicators and discovered six latent factors underlying daily mobility, using GPS data of older adults collected in a study in Germany. To reassure the validity of our framework for selecting representative mobility indicators, we examined the generality and robustness of those six dimensions with another GPS dataset of older adults collected from the MOASIS project. First, we applied the same method to calculate 20 mobility indicators per participant and conduct an exploratory factor analysis (EFA). Second, we ran the EFAs on the mobility indicators of each subgroup of participants by gender, age, and mobility levels. The six dimensions reappeared with minor variations in the mobility indicators of both the entire group and all the subgroups of participants, which implies they are general and robust.