Revisiting the past:replicating fifty-year-old flow analysis using contemporary taxi flow data

Over sixty years ago, geography began its so-called quantitative revolution, where for the first time statistical methods were used to explain the spatial nature of geographic phenomena. Computers made some of this possible, but their limited power did not allow for more than relatively small analytic explorations and consequently many of these earlier ideas are now buried in the mists of time. Here we attempt to replicate one of these early analyses using taxi flow data collected in 1962 and originally used by Goddard (1970 Goddard, J. B. 1970. Functional regions within the city centre: A study by factor analysis of taxi flows in Central London. Transactions of the Institute of British Geographers 49:161–82; then at the London School of Economics) to extract functional regions within London's city center. Our experiment attempts to replicate Goddard's methodology on a modern taxi flow data set, acquired through Global Positioning System tracking. We initially expected that our analysis would be directly comparable with Goddard's, potentially providing insights into temporal change in the spatial structure of the city core. Attempts at replicating the original analysis have proved enormously difficult, however, for several reasons, including the many subjective choices made by the researcher in articulating and using the original method and the specific characteristics of contemporary taxi flow data. We therefore opt to replicate Goddard's approach as closely and as logically as possible and to fill in gaps based on statistically informed choices. We have also run the analysis on two spatial scales—Central London and a wider area—to explore how scales of analyses that were beyond the capacities of Goddard's early computations also help to shape our understanding of the results he obtained