Mixed Effects Modelling with R

Statistics mashup


This is the fourth workshop of the statistics series. This workshop will introduce you to linear mixed-effects models and help develop your theoretical understanding and practical skills for running such models in R.

Linear mixed models are powerful and flexible statistical tools that help us understand the world. This is particularly useful in language sciences when our aim is to investigate the influence of one or more factors (e.g., age; time spent on the Internet per week) on a particular linguistic phenomenon that we are interested in (e.g., speakers' tendency to use internet slang in face-to-face communication). In these cases, we often need to account for some random effects, such as the subtle differences between the different people who produce the language data and/or between the linguistic contexts where the language data is produced. Linear mixed models allow us to do this and more!

This workshop includes two parts. Part 1 is mainly conceptual in which we will discuss what the mixed-effects modelling does and why we want to use mixed-effects models. Part 2 is more practical and we will go through a worked example in R demonstrating how to conduct linear mixed-effect analyses using the lme4 package (Bates, Mächler, Bolker, and Walker, 2015). By the end of the workshop, you will understand the basics of:

· Fixed and random effects

· Random intercepts and slopes

· Model fitting with different random-effects structures

· Model evaluation and selection

This workshop is an advanced training session that requires a basic understanding of R and statistical analyses. This assumed understanding is at the level covered in the previous workshops of the statistics series including “Introduction to Statistical Analysis in R” (11/02) and “Regression with R” (14/02). If you are new to statistics in R we recommend the attendance of these sessions before signing up for this session. The lme4 package should be downloaded and installed prior to participating in this workshop.

The workshop will ultimately build on an understanding of basic statistical analysis, and show how more complex statistical analyses and simulations can be used in R. The relatively advanced techniques and theory behind all of this are highly applicable to wider concepts and will allow you to apply and build on the core concepts with your own research.

After taking part in this event, you may decide that you need some further help in applying what you have learnt to your research. If so, you can book a Data Surgery meeting with one of our training fellows. 

More details about Data Surgeries. 

Those who have registered to take part will receive an email with full details and a link to join the session in advance of the start time. 

If you’re new to this training event format, or to CDCS training events in general, read more on what to expect from CDCS training. Here you will also find details of our cancellation and no-show policy, which applies to this event. 

Return to the Training Homepage to see other available events.