Interpreting Probability Models: Logit, Probit, and Other Generalized Linear Models (Quantitative Applications in the Social Sciences)

"The author provides a stepwise approach for evaluating the results of fitting probability models to data as the focus for the book . . . . All this is packaged very systematically . . . . the booklet is highly successful in showing how probability models can be interpreted."

--Technometrics

"Tim Futing Liao’s Interpreting Probability Models. . . is an advanced text . . . . Liao’s text is more theoretical, but is well exemplified using case studies . . . . this is a text for the more advanced statistician or the political scientist with strong leanings in this direction!"

--John G. Taylor in Technology and Political Science

What is the probability that something will occur, and how is that probability altered by a change in some independent variable? Aimed at answering these questions, Liao introduces a systematic way for interpreting a variety of probability models commonly used by social scientists. Since much of what social scientists study are measured in noncontinuous ways and thus cannot be analyzed using a classical regression model, it is necessary for scientists to model the likelihood (or probability) that an event will occur. This book explores these models by reviewing each probability model and by presenting a systematic way for interpreting results. Beginning with a review of the generalized linear model, the book covers binary logit and probit models, sequential logit and probit models, ordinal logit and probit models, multinomial logit models, conditional logit models, and Poisson regression models.