While the Euclidean method is hypothesis free, the Riemannian framework used here is that the trajectories are to be described in terms of a linear geodesic progression from an initial stage to a final stage, with possible excursions from this linear progression extending over different age ranges and in different directions. This framework is chosen because of the mathematical convenience to start with a linear progression, particularly as it makes comparison between different measures more comprehensible if they are placed in a common framework. It should be emphasized that this is a purely exploratory data analysis; no particular hypotheses are advanced about the structure or content of the developmental process. All statistical analysis is post hoc, devoted to the further understanding of features observed in the initial phase of the analysis.
