Stefan Th. Gries
Home
Contact information
Disclaimer
Last updated: 01 Dec 2020

Teaching at the University of California, Santa Barbara


Ling 204: Statistical methodology (W2020)

Syllabus and overview


This course is a more advanced course on statistical modeling with an emphasis on more sophisticated aspects of regression modeling and other multivariate methods; it presupposes a good understanding of the second edition (2013) of my Statistics for Linguistics with R: [...]. We begin with a first recap of linear and generalized linear regression modeling. We then discuss the use of contrasts and general linear hypothesis tests for linear and generalized linear regression models, followed by some ideas on how to explore curvature in data (regressions with breakpoints, polynomial regressions, and generalized additive models). This is followed by a larger chunk on linear and generalized linear mixed-effects (or multilevel) modeling, where we reanalyze published data and discuss numerical and visual exploration of regression results. The last parts are then devoted to influential data points and validation approaches as well as classification/regression trees and random forests. We use the open source software tool R .


Course downloads



Files for session 01
Files for session 02
Files for session 03
Files for session 04
Files for session 05
Files for session 06
Files for session 07
Files for session 09
Files for session 10

Readings


Software



R from CRAN (make sure you have at least version 3.6.1)
RStudio (make sure you have at least version 1.2.5019)
LibreOffice (make sure you have at least version 6.3.3.2)