Skip to contents

The hmetad package is the most recent implementation of the meta-d’ model. However, this model has had several implementations since its creation.

The first implementation of the model used maximum likelihood estimation for single participant data (Maniscalco and Lau 2012), and is still available for download here. The original code is written in MATLAB, however an updated version is also available in Python.

The model was later implemented by (Fleming 2017) in a hierarchical Bayesian framework, which has been shown to provide much more reliable estimates in the relatively small sample sizes commonly used in psychological experiments. This version, known as the Hmeta-d toolbox, was implemented in the probabilistic programming language JAGS, which in turn has interfaces in both MATLAB and in R.

The hmetad package builds on these previous versions through implementation in the brms package in R, allowing for estimation with complex regression designs. Additionally, brms uses the probabilistic programming language Stan, which permits much more efficient sampling with more reliable model convergence warnings. Because of its increased efficiency and flexibility, the hmetad package is now the recommended approach to fitting the meta-d’ model.

References

Fleming, Stephen M. 2017. “HMeta-d: Hierarchical Bayesian Estimation of Metacognitive Efficiency from Confidence Ratings.” Neuroscience of Consciousness 2017 (1): nix007.
Maniscalco, Brian, and Hakwan Lau. 2012. “A Signal Detection Theoretic Approach for Estimating Metacognitive Sensitivity from Confidence Ratings.” Consciousness and Cognition 21 (1): 422–30.