Researchers in the social sciences often require reference books to aid them in the computation and interpretation of statistics. These books are usually organized around a set of statistical tools and give extensive detail to the formulas used in estimation. Researchers also use computer packages to compute these statistics. A simple, oneline command will give birth to multiple regression, factor analysis, or analysis of  variance. On the bookshelves next to the statistical texts and computer manuals are the books that summarize and interpret the field. These texts of substantive theory may occasionally present statistical
summaries of data, but data play a small role in such texts. To some extent, theory and data rarely touch each other in the social sciences. To bridge the gap, over the past ten or so years, an area called causal or
structural analysis has developed. This area takes as its task the explication of a statistical method in terms of how the method relates to substantive theory.
David A. Kenny

在国外社会科学研究方面好像很有影响的一本书，尽管比较老，但引用仍比较多，但愿有用。