Interpreting Economic and Social Data, by Othmar W. Winkler, Berlin, Springer-Verlag, 2009, xvi+265 pp., £81.00 or US$119.00 (hardback), ISBN 978-3-540-68720-7
The book presents a critique of the conventional application of statistics in social sciences as a watered-down version of mathematical statistics. It therefore introduces the concept of “real-life objects” as “statistical counting units” based on the reasoning that socio-economic parameters fundamentally differ from those generated by conventional statistical measurements in natural sciences which form the basis of statistics. Its main theme revolves around the aggregation of the objects in a three-dimensional space of subject matter, time period and geographical coverage. More specifically, it calls for more emphasis on the interpretation of results than on the computing aspect of socio-economic data. The rationale for such an emphasis is that it helps avert the potential risks of assigning more weight to the “significance/non-significance” of the results than to their interpretations based on expert knowledge. More generally, the book calls for a unified approach towards understanding the mathematical and statistical aspects of society.
The book is divided into 12 chapters forming three main clusters: understanding the socio-economic structure and data; longitudinal and cross-sectional analysis of socio-economic data and the relation between socio-economic statistics with probability, accounting and geography. It is written in a coherent style with no requirement of mathematical maturity to follow and understand. While it is not suitable as an introductory statistics text, it is clearly useful to researchers in the social sciences – particularly to those seeking to uncover and understand socio-economic structure through data analysis. At first glance, multi-disciplinary researchers may detect potential gaps between the book's philosophy and the recent developments in multi-disciplinary research. However, it is imperative to believe that the book is also reiterating the multi-disciplinary approach to data analysis albeit from a slightly different perspective. Given its positive approach, I believe future versions of the book have the potential to broaden and supplement these kinds of understanding.
There are a few typographical errors across the chapters. Notes are given at the end of each chapter in the form of end notes, a style which may be rather disruptive as the reader is made to move forth and back – quite often through huge chunks of text. Weighed against the depth of additional details provided in the notes, the disruption is probably negligible.