ABSTRACT
This chapter examines some extensions of basic correlational methods. The first topic is factor analysis, a set of techniques for determining the extent to which a set of variables represents one or more underlying dimensions. Exploratory factor analysis starts with a set of variables and searches for relationships among them, with the goal of identifying one or more dimensions or factors underlying the set of variables being analyzed. Confirmatory factor analysis (CFA) starts with a hypothesis about the pattern of relationships among a set of variables and tests that hypothesis; it also can be used to assess the structural validity and generalizability of measures. Indices of goodness-of-fit are used to evaluate whether the results of the analysis match the hypothesized factor structure. The chapter provides an overview of methods for testing mediational hypotheses (or models), which propose that one or more variables come between an independent variable and a dependent variable in a causal sequence. These methods include the causal steps strategy, path analysis with observed variables, and structural equation modeling. Prospective research designs are also discussed, as are limitations on interpreting the results of mediational analyses.
