Psychology 6136: Categorical Data Analysis

• Instructor: Michael Friendly (my home page)
  • Email: friendly AT yorku DOT ca
  • Office: 226 BSB
  • Phone: x66249
  • Office hour: Wednesday: 11:30-12:30 (other times by appt.)
• Class meetings: Thursday, 2:30 am - 5:30 pm, 204 BSB, lab session: 4:30-5:30, Hebb lab, 159 BSB. The first class will be on Sep. 7.

Text books and readings

Main texts

The main texts for this course are

• Friendly, M. and Meyer, D. (2016). Discrete Data Analysis with R: Visualization and Modeling Techniques for Categorical and Count Data . Chapman & Hall. ISBN 9781498725835. Web site for book: ddar.datavis.ca
• Agresti, A. (2007). Introduction to Categorical Data Analysis, 2nd ed., NY: Wiley. ISBN 9780470114742.

Supplementary readings:

• Agresti, A. (2013). Categorical Data Analysis, 3rd ed., NY: Wiley. A much more technical book, that many consider the "bible" for categorical data analysis methods. There is also a manual for R and S-plus users to accompany this text.
• Fox, John. Applied Regression Analysis and Generalized Linear Models, 2nd Ed. Sage, 2008. An excellent text on linear models; Part IV on Generalized Linear Models provides a clear and comprehensive discussion.
• Grolemund & Wickham, R for Data Science A good tutorial book on the tidyverse approach to data analysis in R. Available free online.
• Fox & Weisberg An R Companion to Applied Regression, 2nd Ed., Sage, 2011. There is also a web page for the book, containing data files, R scripts and a collection of web appendices on other topics.

Topic schedule and lecture notes

Week	Topic	Readings	R files
1	Overview [slides] [4up]	DDAR: Ch1, Ch2; Agresti: Ch1	R-into.R []
2	Discrete distributions [slides] [4up]	DDAR: Ch3	R-data.R [] binomial.R []
3	Two-Way Tables: Independence and Association [slides] [4up]	DDAR: Ch4; Agresti: Ch2	berk-4fold.R [] vision-sieve.R []
4	Two-Way Tables: Ordinal Data and Dependent Samples [Tutorial] on two-way tables	DDAR: Ch4; Agresti: Ch2	msdiag-agree.R [] haireye-spineplot.R []
5	Loglinear Models and Mosaic Displays [slides] [4up] [Tutorial] on loglin models; [Mosaic display animation]	DDAR: Ch5; Agresti: 2.7, Ch. 7	berkeley-glm.R [] titanic-loglin.R []
6	Correspondence Analysis [slides] [4up]	DDAR: Ch6	mental-ca.R [] mca-presex3.R []
7	Logistic Regression I [slides] [4up] [Logistic regression tutorial]	DDAR: 7.1-7.3; Agresti: 3.1-3.2; Ch 4	arthritis-logistic.R [] cowles-logistic.R [] Arrests-logistic.R []
8	Logistic Regression II [slides] [4up]	DDAR: 7.3-7.4; Agresti: Ch 4-5	cowles-effect.R [] Arrests-effects.R [] berkeley-diag.R []
9	Multinomial Logistic Regression [slides] [4up]	DDAR: 7.5-7.6; Agresti: Ch 6	arthritis-propodds.R [] wlf-nested.R [] wlf-glogit.R []
10	Log-Linear Models I [slides] [4up]	DDAR: 8.1-8.4; Agresti: Ch 7
11	Log-Linear Models II [slides] [4up]	DDAR: 8.5-8.11; Agresti: Ch 8
12	Generalized Linear Models: Count Data[slides] [4up] [Count data GLMs tutorial]	DDAR: 9.1-9.4; Agresti 3.3-3.5	phdpubs.R [] quine.R []
13	Generalized Linear Models: Further topics	DDAR: 9.5-9.7

Assignments

Please submit your assignments to me by email, as a PDF, Word, or HTML attachment (together with the associated R file), with a Subject: line "PSYC 6136: Assignment XX". To help me keep them straight, it would be most convenient to name them something like "YourName-AssignXX.{pdf,docx,html}".

• Assignment 1 [R solutions ]
• Assignment 2 [R solutions ]
• Assignment 3 [R solutions ]
• Assignment 4

Evaluation

There are three components to your evaluation in the course: two take-home projects (each worth 40%) that will involve analysis of one or more data sets together with a research report describing the background, your analyses, results and conclusions. For these, you can use any software you like, although R is strongly encouraged.

Here is a template for a markdown .Rmd file you can use if you are working in RStudio and want to write in R markdown.

1. Project 1 : a selection of data sets for the material up to and including logistic regression. Due date: Nov. 3

2. Project 2: a selection of data sets for the material from logistic regression to the end of the course. Due date: Dec. 15

3. The remaining 20% can be earned either as
   1. an assignment portfolio, containing a selection of your best work (possibly edited/enhanced) on a selection of assignment questions, or by
   2. reading and discussing a journal article related to theory or application of categorical data analysis. For the latter, you can volunteer to give a brief (~15 min) presentation to the class (sometime in Nov.) to earn bonus marks. Due date: Dec. 29

Resources

Statistical software

• R Project for Statistical Computing. R is a free software environment for statistical computing and graphics (Windows, MacOS, Linux).
  You can download it from any CRAN mirror site.
  See John Fox's notes, Installing R for Windows. Somewhat out of date, but still useful.
• For this course, use this R script to install useful add-on packages for categorical data analysis.
• R Studio is a powerful front-end for R, much more convenient than the standard R GUI. Among other features, it offers integrated tools for help, plotting, history and the ability to run an R script to obtain a HTML, PDF or Word document containing your input and output.
• If you insist on SAS or SPSS, see York group licenses for SAS and SPSS
  You may also be able to use the Web Acadlabs service to access SAS, SPSS from home. See
• Accessing software remotely using Web Acadlabs by Manolo Romero.