The literature_review.txt file contains the full list of references for our literature review, and our reasoning behind selecting papers for steps 1), 2) and 3) of our paper selection method.
The code folder contains the code and data (master_file.json,master_file.csv) we used in our analyses.
graph_analysis.me.R includes our mixed effects models for max depth, max breadth, and aspect ratios.
interaction_rates.R includes our mixed effects models for interaction rates.
task_answers.csv provides details about participants' task completions and the types of answers they provided.
prompt_states.json includes the data attributes associated with achieving task goals.
Our overlap (histograms, modified Jaccard similarity) and loop analyses are computed using graphAnalysis.py, overlap.py, and getBaseStats.py
think_time.csv, think_time_stats.csv, think_time_sequences.csv and open_ended_inputs.csv include the data for our timing analyses. Our analyses are performed using thinkTime.py, thinkTimeSequences.py, think_time.R and think_time_sequences.R.