This OSF project accompanies the manuscript:
Spektor, M. S.\*, Kellen, D.\*, Rieskamp, J., & Klauer, K. C. (2024). Absolute and relative stability of loss aversion across contexts. *Journal of Experimental Psychology: General*, *153*(2), 454-472. https://doi.org/10.1037/xge0001513
The project is organized as follows:
1. **main_analyses.ipynb**: A Jupyter (https://jupyter.org/) notebook that runs the main analyses reported in the manuscript and generates the behavioral figures
2. **order_constrained.R**: An R (https://www.r-project.org/) script containing the order-constrained inferences reported in the main text
3. **model_fitting_all_data.ipynb** and **model_fitting_shared.ipynb**: Jupyter (https://jupyter.org/) notebooks containing the model-fitting procedures for the streamlined cumulative prospect theory model, once fitted to all trials and once fitted to trials that were shared across the two conditions
4. **posterior_predictives.ipynb**: A Jupyter (https://jupyter.org/) notebook that uses the model fits created by **model_fitting_all_data.ipynb** to create the posterior predictive figures in the Appendix
5. The folder **data** contains the anonymized data from each of the experiments
6. The folder **figures** contains all figures created by **main_analyses.ipynb** and **posterior_predictives.ipynb**
7. The folders **fits_all_data** (model fitted to all trials) and **fits_shared** (model fitted to trials that were shared across the two conditions) contain for each experiment (**exp1** to **exp3**):
- **_badfirst_MV.nc**: Fitted posterior distribution of the hierarchical multivariate cumulative prospect theory for participants who started in the gain-seeking condition first. File is in the NetCDF format and can be opened, for example, using ArviZ (https://python.arviz.org)
- **_goodfirst_MV.nc**: Fitted posterior distribution of the hierarchical multivariate cumulative prospect theory for participants who started in the loss-aversion condition first. File is in the NetCDF format and can be opened, for example, using ArviZ (https://python.arviz.org)
- **_MV.nc**: Fitted posterior distribution of the hierarchical multivariate cumulative prospect theory for all participants. File is in the NetCDF format and can be opened, for example, using ArviZ (https://python.arviz.org)
8. The folder **models** contains the **fullCPTMV.stan**, the model file for Stan (https://mc-stan.org/) for the model used for the parametric analyses
9. The folder **data** contains the experimental trials from each of the experiments
10. The linked project **Experiment 3** contains the time-stamped preregistration documents (preregistration, blinding procedure, blinded and unblinded data, experimental code) for Experiment 3.
Correspondence should be sent to: Mikhail Spektor ([mikhail@spektor.ch][1]) or David Kellen ([davekellen@gmail.com][2])
[1]: mailto:mikhail@spektor.ch
[2]: mailto:davekellen@gmail.com
