notebooks

This folder contains notebooks to replicate analyses and figures from the paper and supplement. Notebooks are generally ordered according to when analyses are described or figures appear in the paper.

main/

code to reproduce analyses and figures from the main text

• 1_model-course-content.ipynb – code to fit the topic model, transform lectures & quiz questions, and generate Figure 2C
• 2_topic-variability-fig.ipynb – code to generate Figure 3 and compute related stats
• 3_lecture-question-timeseries-fig.ipynb – code to generate Figure 4
• 4_reconstructing-knowledge.ipynb – code to estimate dynamic knowledge traces, generate Figure 5, and compute related stats
• 5_predictive-analyses.ipynb – code to generate Figure 6
• 6_knowledge-smoothness.ipynb – code to generate Figure 7
• 7_knowledge-maps.ipynb – code to construct knowledge maps and generate Figure 8
• 8_parse_demographics.ipynb – code to compile demographic data described in Methods
• 9_get-transcripts.ipynb – code to fetch automated transcripts from the YouTube API as described in Methods

supp/

code to reproduce tables and figures from the Supplementary Materials document

• 1_tables.ipynb – code to generate Supplementary Tables 1–3 and compute the correlation timeseries peaks shown in Supplementary Figures 3 & 4
• 2_topic-word-distributions-fig.ipynb – code to generate Supplementary Figure 1
• 3_topic-weight-fig.ipynb – code to generate Supplementary Figure 2 and compute related stats
• 4_lecture-question-correlation-peaks.ipynb – code to generate Supplementary Figures 3 & 4
• 5_model-comparison-word-overlap.ipynb – code to generate Supplementary Figure 5
• 6_model-comparison-BERT.ipynb – code to generate Supplementary Figure 6
• 7_individual-maps.ipynb – code to generate Supplementary Figures 7–11