Applied regression allows social scientists who are not specialists in quantitative techniques to arrive at clear verbal explanations of their numerical results. Provides a lucid discussion of more specialized subjects: analysis of residuals, interaction effects, specification error, multicollinearity, standardized coefficients, and dummy variables.