This is the online supplement to the manuscript "**Cognition-Enhanced Machine Learning for Better Predictions with Limited Data**" submitted to *Topics in Cognitive Science*.
The necessary [data has been made available on Dataverse by the Duolingo researchers][1] and can be downladed there. For more information, [see their description][2].
Our analyses produced two large CSV files called `exp_df_train.csv` and `exp_df_test.csv` that contain the models' fits on the training data and their predictions for the test data, respectively. Additionally, the `Results Data` folder contains results for the analyses referred to as "slices of the dataset" in the manuscript.
The two RMarkdown notebooks that are included implement all the analyses and create all numbers and figures included in the manuscript. Specifically, `results_figures.Rmd` includes the analyses while `duo_data_overview.Rmd` provides general descriptives and many additional visualizations that we hope give the reader a better impression of the data used for the reported analyses.
[1]: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/8SWHNO
[2]: https://sharedtask.duolingo.com/2018.html