This repository implements the Global Matching Model (GMM) from Osth & Dennis (2015) to explore the test position effect in recognition memory.
runme.R: Runs model simulations. This is the main script to generate predictions. You can set parameter values and task design features here (e.g., list length, task type).model.R: Computes the means and variances of the old and new item distributions based on item- and context-level match/mismatch.tpe.R: Estimates changes in accuracy over the course of testing (i.e., the test position effect) for yes-no recognition and alternative forced-choice tasks.afc.R: Contains a function to compute accuracy for an Alternative Forced Choice (AFC) task using parameters from normal distributions.
To generate predictions, run the runme.R script after adjusting the parameters and task design as needed.
For theoretical background on the model, please refer to:
- Osth & Dennis (2015). Psychological Review.
Feel free to reach out with any questions: aytac.sinem@gmail.com