CMStatistics 2022: Start Registration
View Submission - CMStatistics
B1716
Title: Spatio-temporal Bayesian methods for historical data Authors:  Tiia-Maria Pasanen - University of Jyväskylä (Finland) [presenting]
Abstract: A Bayesian spatio-temporal model is developed to study pre-industrial grain market integration during the Finnish famine of the 1860s. The data consists of 80 regional time series covering nine years of monthly grain prices. The model takes into account several problematic features often present when analysing such spatially interdependent time series. For example, compared with the error correction methodology commonly applied in econometrics, this approach allows simultaneous modelling of multiple interdependent time series, avoiding cumbersome statistical testing needed to predetermine the (often artificial) market leader as a point of reference. Furthermore, introducing a flexible spatio-temporal structure enables analysing of detailed regional and temporal dynamics of the market mechanisms, for example, the asymmetric neighbour dependencies. The whole process, from deriving the model to interpreting the results, is covered with the famine application from the point of view of a statistician.