Categorical Data Analysis (8 ECTS)

Categorical data types. Contingency tables, joint, marginal and conditional probabilities, independency, comparing rates in 2x2 contingency tables (2 rate difference, relative risk, relative probability ratio), types of observant research (recursive, cross-sectional, perspective), relative probability and other correlation measures in IxJ matrices. X² test of independence, exact tests, X² statistical test partition, independence test for regular data, trend tests for 2xJ tables. Associated data pairs, comparing associated rates, Mc Nemar test for comparing marginal rates, measures of agreement between observers, relative probability for agreement rate, kappa measure of agreement. Correlation in multidimensional contingency tables, conditional and marginal relative probability rates, the Simpson paradox, partial-conditional independence, homogenized correlation, collapsibility, Cochran-Mantel-Haenszel tests. Logistic regression, model parameters interpretation, logistic regression inference, the case of categorical predictive variables, multiple logistic regression, model choice, model sufficiency test.

     Recommended Reading

• Agresti A., (2013). Categorical data analysis, Wiley
• Agresti A., (2007). An Introduction to Categorical Data Analysis, Wiley.
• Hosmer, D., Lemeshow, S. and Sturdivant, R. (2013) Applied Logistic Regression, Wiley
• Kateri, M. (2014). Contingency Table Analysis, Springer.