Exceptions Are the Rule - pocket, Engelsk, 1993
How do you relate concepts like religion and politics in numbers and equations? This book attempts to do just that by making quantitative social science easier. The author's essays aim to uncover the hidden links between theory and method and disclose the pitfalls along the way as he lead the reader through philosophy and realism to a clearer understanding of social science. Levine shows why quantitative social science must develop by its own rules, distinct from those that govern the disciplines of mathematics and statistics. Throughout, he emphasizes the use, misuse and meaning of common techniques, and the extension of basic tools to new problems. The book explains how to let measurements emerge during the analysis of data in ways that minimize a priori assumptions while maximizing the information extracted from data. For scholars, the book builds on debates between methodologists and theoreticians about the future of statistical methodology, presenting a unified analysis of diverse applications in economic time series, political identity, small group sociometry and occupational stratification.