ABSTRACT
This book was written especially for new students of social science. Presumably such students gradually achieve greater levels of sophistication in their understanding of regression analysis and other forms of multivariate analysis as their training in statistics and research methods progresses. The book helps students to realize that even before they study statistics, they can learn enough about multivariate analysis to understand at a basic level-and even to critically evaluate-empirical research relying on regression or other multivariate techniques. The most important requirement for such research is that it be based on sound theory. The second most important requirement for good multivariate analysis is that the model accurately specify the theory being tested, that is, it should include all independent variables central to the theory and reflect any forms of interaction and nonlinearity hypothesized.
