/cmips

Computational Methods Investigating Psychosocial Stressors (CMIPS)

Primary LanguagePythonGNU General Public License v3.0GPL-3.0

CMIPS Project

Computational Methods Investigating Psychosocial Stressors (CMIPS).

Code used in the doctoral dissertation of Cory J. Cascalheira to predict psychosocial stressors from the social media of sexual and gender minority adults.

Using The Code

All code is free to use with attribution to the author, Cory J. Cascalheira.

If you use the bot detection tactics (BDTs) in your work, please cite the corresponding paper: https://psyarxiv.com/gtp6z/

Order of Script Exection

Clean and Preprocess, Part I 0. src/recruitment_sources.R

  1. src/bdt_script.R
  2. src/clean/clean_qualtrics.R
  3. src/clean/clean_strain.R
  4. src/clean/combine_dichotomize.R
  5. src/extract/*
  6. src/clean/combine_social_media.R

Analyze, Part I 7. src/analyze/describe_social_media.R

Clean and Preprocess, Part II 8. src/clean/clean_social_media.py 9. src/clean/clean_social_media.R

Create Features / NLP-Generated Independent Variables 10. src/create_features/create_features_* 11. src/create_features/create_features_dassp/* 12. src/create_features/synthesize_features.R

Merge Features and Survey Data 13. src/combine_features_scores.R

Analyze, Part II 14. src/analyze/describe_social_media.R 15. src/analyze/describe_participants.R 16. src/analyze/describe_measures.R 17. src/analyze/ml_modeling/classify_.py 18. src/analyze/ml_modeling/regress_.py 19. src/analyze/feature_selection/pca_* 20. src/analyze/feature_selection/feature_selection.py 21. src/analyze/feature_selection/top_features.R 22. src/analyze/features_survey_predictors/*