The Data Game: Controversies in Social Science Statistics

Summary: While not a computer book, this work discusses data that are fundamental to much of what social scientists do with computers. Maier discusses various data sources, such as the census, the American Housing Survey, the Uniform Crime Reports, Federal Reserve Board statistics, and others, presenting the controversies behind each. For instance, with regard to the census, Maier discusses population undercounting, undocumented immigrants, difficulties of predicting birth rates, problems in interpreting the ratio of elderly to working population, issues about who is Hispanic or black, the difference made by considering households rather than families, state differences in divorce statistics, and many other issues. The points are made in a lively, concise manner that goes a long way toward illuminating the social science research process as well as its pitfalls. For instance, the section "Oh No, I Forgot to Get Married" discusses the data and forecasting assumptions behind the much-publicized social science report of 1986 that asserted that women who had not married by age 35 had only a 5% chance of ever marrying. By examining dozens of cases of this type Maier has thrown light on