Department of Sociology—Sociological Research Methods Cluster

Short Course: Categorical Data Analysis for Social Scientists

June 20/21 2016

Instructor: Dr Brendan Halpin

Applications are invited for a two-day workshop on Categorical Data Analysis for Social Scientists, June 20 and 21, 2016.

The course is aimed at social science researchers (PhD students, academics and others) who have a grounding in quantitative methods and wish to improve their categorical data analysis skills. The course will combine theory and practice, with hands-on sessions using Stata.

The key topics covered will include:

• association in nominal and ordinal categorical data: tests and measures
• logistic regression: regression models for binary outcomes
• multinomial logistic regression: regression for multiple-category outcomes
• ordinal logistic regression: regression for ordered outcomes
• models for count data: poisson regression and related models, loglinear models
• marginal effects in non-linear models

The course will draw on Long and Freese (2014), Regression Models for Categorical Dependent Variables Using Stata, and Agresti (2007), Introduction to Categorical Data Analysis.

Participants should have a good grounding in basic statistics, up to and including a working knowledge of multiple regression. Experience with statistical software (not necessarily Stata) will be an advantage.

Class notes

Class notes for the course are available in these formats:

• PDF, 1 slide per page
• PDF, 4/page
• PDF, 8/page
• HTML

Lab notes

Lab notes in online and printable format.

Locations

Registration and classes will be held in F1030, Foundation Building, UL.

PC Labs will take place in A0060a, Main Building.