Skip to main content
Languages, Linguistics and Film

PGR Training: Mixed Effects Modelling

When: Wednesday, November 13, 2024, 2:00 PM - 4:00 PM
Where: Bancroft 3.17, Mile End Campus

Part of our PGR training programme, this event is open to PGR students only, including MA, MRes and PhD students. 

Please click 'book now' to sign up in advance.

Training leader: Matt Hunt Gardner


Logistic regression is considered the "gold standard" for variationist sociolinguistic research (Tagliamonte 2012, etc), but often the idiosyncrasies of actually doing logistic regression in R (the current go-to statistical modelling program) can create headaches for linguists less familiar with statistics. This workshop will focus on demystifying mixed-effects logistic regression as it is implemented by the lme4 package in R. Topics covered will include a basic theoretical explanation of what exactly logistic regression is, the difference between fixed effects, random effects, and random slopes, the relationship between older statistical tools (i.e., Goldvarb) and the lme4 package in R, common pitfalls to avoid, and tips and tricks to making doing and reporting on a logistic regression analysis easier.  If you ever plan on publishing a manuscript that includes logistic regression, this workshop will help you head off reviewer concerns before you even submit. 

A general awareness of sociolinguistic methods is assumed, but those from other disciplines looking to use logistic regression to model binary variables will find this workshop useful. Prior experience/knowledge of R is not essential.  
A step-by-step explanation of data wrangling (an essential component of preparing to do logistic regression) will not be covered, but can be practiced using the tutorial "Doing LVC with R" at https://lingmethodshub.github.io/content/R/lvc_r (Gardner 2020). 
You do not need to bring a laptop for this workshop, but do please bring any questions you have about mixed-effects logistic regression or statistical methods used in variationist sociolinguistics. Summary handouts will be distributed. 
If you have any questions about the workshop, please direct them to Matt: matt.gardner@qmul.ac.uk

Back to top