Abstract
Causal modeling has become an increasingly popular heuristic and analytical tool in the social sciences. My emphasis in this monograph is on those causal analysis techniques most appropriate for cross-sectional, interval-level data, although some discussion of the causal analysis of two- and three-wave panel data will be presented. The techniques to be discussed range from correlational tests for spurious relationships in simple three-variable models to estimation procedures for complex models with nonrecursive (reciprocal) linkages. The techniques presented herein are not appropriate for time-series data consisting of observations on the same variable collected at multiple (normally ten or more) time points; the analyst with such data should examine the monographs by Ostrom (1978) and by McDowall et al. (1980) in this series. Likewise, if one ...