A Stan reimplementation of the knower-level model

Refer to this blog post for a detailed introduction to the background and implementation.

Data

• Raw data from Bayesian Cognitive Modeling (Lee & Wagenmakers, 2013): raw/NumberConcepts/fc_given.mat
• Exported JSON format: made/data.json

Source Code

• Model source code: src/m2.stan
  • src/m3.stan: Reference implementation for the original knower-level model. It is a syntactic update of Martin Smira's Stan translation of the original WinBUGS code in the book.
• Validation of model on synthetic data: src/sim.R, src/validate.R
• Fitting model to the real Give-N data: src/fit2.R

Inference Results

The results are NOT identical to those presented in Lee & Wagenmakers (2013). Refer to the blog post for details.

References

• Lee, M. D., & Sarnecka, B. W. (2010). A Model of Knower-Level Behavior in Number-Concept Development. Cognitive Science, 34(1), 51–67.
• Lee, M. D., & Sarnecka, B. W. (2011). Number-knower levels in young children: Insights from Bayesian modeling. Cognition, 120(3), 391–402.
• Lee, M. D., & Wagenmakers, E.-J. (2013). Bayesian cognitive modeling: A practical course. (pp. xiii, 264). Cambridge University Press.