tangrams

This repository contains the experiment, data, and analysis code for the paper "Characterizing the dynamics of learning in repeated reference games".

Manuscript pre-print available on ArXiv

• experiments contains the code and materials for running repeated reference games as web experiments
• data contains the raw, anonymized data from the experiments as well as the pre-processed versions (unconstrained, sequential)
• analysis contains the python and R code for producing all results and figures reported in the paper
• writing contains the TeX to generate the journal manuscript