mixtur

mixtur is an R package for designing, analysing, and modelling continuous report visual short-term memory studies. The package allows users to implement the 2-component (Zhang & Luck, 2008) and 3-component (Bays, Catalao, & Husain, 2009) mixture models of continuous-report visual short-term memory data. The package can also fit & simulate the slots and slots-plus averaging models of Zhang & Luck.

The package allows users to:

• Obtain summary statistics of response error and response precision of behavioural data
• Produce publication-ready plots of behavioural data
• Fit both the 2- and 3-component models to user data
• Plot the goodness of model fit to user data
• Simulate artificial data from both models
• Conduct formal model competition analysis

Pre-print

We have a pre-print showing users how to use the package:

• Grange, J.A. & Moore, S.B.(under review). mixtur: An R package for designing, analysing, and modelling continuous report visual short-term memory studies. https://psyarxiv.com/n6gqx/

The pre-print also includes several simulation studies exploring some properties of the models (including parameter recovery simulations, model recovery simulations) and provide concrete recommendations to researchers wishing to use mixture modelling in their own research.

Installation

To install the package from GitHub, you need the remotes package:

install.packages("remotes")

Then mixtur can be directly installed:

remotes::install_github("JimGrange/mixtur")
library(mixtur)

Acknowledgements & References

• We are grateful to Ed. D.J. Berry who contributed to the package development.

• Portions of the package code have been adapted from code written by Paul Bays in Matlab, with permission. We are extremely grateful to Paul Bays for this permission. See https://paulbays.com.

• Bays, P. M., Catalao, R. F. G., & Husain, M. (2009). The precision of visual working memory is set by allocation of a shared resource. Journal of Vision, 9(10): 7, 1–11.

• Zhang, W., & Luck, S. J. (2008). Discrete fixed-resolution representations in visual working memory. Nature, 453, 233–235.