Note: corresponding datasets and results are in private folders, access via request.
Run script: python run.py fa23-config-copy.py data/responses_clean.csv
Original data from ./data/responses.csv
Outputs from running the algorithm ./out-private-output.csv
Post processed result ./out-private_post_process_FINAL.csv
Repo dedicated to hosting all code and files associated with grouping students into groups based on configurable constraints and CSV format.
With Python 3.x installed, run:
python run.py configs/group_matching_template.py /path/to/matching_form_responses.csv
"...group_matching.pdf" -- Initial group matching surveys "...evaluation_survey.pdf" -- Final group feedback surveys
The surveys with the prefix "16A_Fa20..." were used in the first iteration of our group matching system.
Changes from Fall 2020:
- Asked for preferred discussion section times in intake survey
Changes from Fall 2020:
- Asked for preferred discussion section times in intake survey
- Asked whether the student would be on campus or not following COVID-19 restrictions
Changes from Fall 2020:
- Asked for preferred discussion section times in intake survey
- Asked whether or not the student would like to lead their group
- Asked how the student would like to interact with their study group
- Asked whether or not the student prefers a study group where another student shares a common identity with them
Changes in group matching for mandatory project groups in CS 169A ("CS169_Fa22..."):
- Groups were restricted to a group size of exaclty 4
- Final evaluation survey asked Physcology Safety questions on a seven-point Likert scale
Changes from Fall 2020:
- Asked for preferred discussion section times in intake survey
- Asked whether or not the student would like to lead their group
- Asked how the student would like to interact with their study group
- Asked whether or not the student prefers a study group where another student shares a common identity with them
- The evaluation survey did not change between Spring 2022 and Spring 2023.
- Surveys varied slightly in different classes in the same semester, but only one class's survey for a given semester is represented in this repository. This survey is an approximation of all of the surveys released for that semester.