Unifying Political Methodology: The Likelihood Theory of Statistical Inference

by Gary King

In this book the author demonstrates the utility of the likelihood theory of statistical inference as a basic, unifying approach to statistical modelling for political research. This approach encourages the tailoring of models to particular phenomena studied.

Genres: Grad School

286 pages, Hardcover

First published August 25, 1989

Reviews

[Hugo Salas · Rating 4 out of 5]

Painful. But enlightening for those interested in Maximum Likelihood methods.

[B Frizz · Rating 2 out of 5]

didn't actually read the entire book (#stats), but talk about a horribly inaccessible statistics book for social scientists...

[Rich · Rating 3 out of 5]

A good book: 3.5/5. But one that you wouldn't want to start with if desiring to learn about likelihood techniques of estimation--but one that can well build and extend the lessons you've already internalized to think about likelihood more broadly than just another technique for analysis. Instead, start with Long (1997). Here, King argues that likelihood is a basis for inference in social science, not just a tool for analysis, and it is a strong argument for political science to question the habits it has developed in borrowing its tools from other areas of research (econometrics) and earlier times (before computers where efficiency in estimation was perfunctory).