Replication Data for “Cognitive Costs in Dynamic Belief Revision: Common Regularities and Persistent Individual Heterogeneity”
Description
This dataset contains the de-identified derived data, analysis code, frozen results, and experimental materials for the article “Cognitive Costs in Dynamic Belief Revision: Common Regularities and Persistent Individual Heterogeneity.” The experiment involved 81 participants completing 5,184 incentivised sequential learning-and-stopping trials, of which 5,159 contained at least one recorded signal pair. The files cover assigned priors, Normal and Bernoulli signals, refresh intervals, discount assignments, stopping times, terminal choices, slider events, Bayesian benchmark paths, and collapsed belief revisions. The Python and MATLAB code reproduces the reported descriptive statistics, regression models, randomisation and bootstrap analyses, condition-matched comparisons, and empirical figures. README files and software requirements document the analysis workflow. Participant identifiers are pseudonyms P01 through P81. Original participant records, identity crosswalks, consent forms, and payment records are not included. The package also contains the original Japanese experimental instructions and an English translation.
Files
Institutions
- Tokyo University of ScienceTokyo, Tokyo