R PACKAGE psycholing
We're putting together an R package with functions for performing standard data processing tasks for many common experimental designs in psycholinguistics and cognitive psychology and for coding and interpreting the results for linear mixed-effects models.
Here's how you can grab and use the current development version of the package:
- Install the devtools package if you haven't already. It's available on CRAN.
- In R, type devtools::install_github('sfraundorf/psycholing') to both download and install the package in one step.
- That's it! You now have the psycholing package installed. You can load it up with library(psycholing) and use it just like any other R package.
You can also check out the GitHub repository for the package.
By popular demand, here's a Web-based version of Scott's course in linear mixed-effect models, updated for the fall 2020 term. Disclaimer: These lecture notes are from a previous term, so there's a possibility they no longer reflect current practice in the field.
- Introduction to multi-level / mixed effects models
- Descriptive statistics in R
- Data processing in R
- Fixed effects
- Fixed-effect interactions and outliers
- Model comparison
- Random intercepts
- Level-2 variables
- Random slopes
- Crossed random effects
- Centering and other transformations
- Contrast coding for variables with two categories & "Main effects" vs. "simple effects"
- Post-hoc comparisons
- Contrast coding for variables with more than two categories & orthogonality
- Logit models for categorical outcomes
- Empirical logit and Poisson models
- Dealing with missing data
- Longitudinal and time-series data: Growth curve analysis
- Longitudinal and time-series data: Autocorrelation
- Crossed-lagged designs and effect size
- Statistical power
- Signal detection theory