Regression With Social Data: Modeling Continuous and Limited Response Variables

Regression with Social Data: Modeling Continuous and Limited Response Variables represents the most complete and fully integrated coverage of regression modeling currently available for graduate-level behavioral science students and practitioners. Covering techniques that span the full spectrum of levels of measurement for both continuous and limited response variables, and using examples taken from such disciplines as sociology, psychology, political science, and public health, the author succeeds in demystifying an academically rigorous subject and making it accessible to a wider audience.

Content includes coverage of:

• Logit, probit, scobit, truncated, and censored regressions
• Multiple regression with ANOVA and ANCOVA models
• Binary and multinomial response models
• Poisson, negative binomial, and other regression models for event-count data
• Survival analysis using multistate, multiepisode, and interval-censored survival models

Concepts are reinforced throughout with numerous chapter problems, exercises, and real data sets. Step-by-step solutions plus an appendix of mathematical tutorials make even complex problems accessible to readers with only moderate math skills. The book’s logical flow, wide applicability, and uniquely comprehensive coverage make it both an ideal text for a variety of graduate course settings and a useful reference for practicing researchers in the field.