Skip to content

In this project, my colleague Catherine Lee (Rutgers) and I employ computational text analysis to examine quantitative trends in the use of diversity terms, OMB/Census terms, and other population labels in a sample of 2.6+ million biomedical abstracts spanning the last 30 years.

License

Notifications You must be signed in to change notification settings

brandonleekramer/diversity

Repository files navigation

The Rise of Diversity and Population Terminology in Biomedical Research

As of: 05-17-2021

This repository provides the source code for the Brandon Kramer and Catherine Lee's "The Rise of Diversity and Population Terminology in Biomedical Research." After uploading the PubMed/MEDLINE database with PubMedPortable in Python, we used R's tidytext package to examine trends in the use of diversity in more than 2.5 million scientific abstracts from 1990-2020. Overall, our analyses demonstrate that various types of "diversity" and other population terminiology, including race and ethnicity, are rising over time. While we provide some prelimiary results and a full appendix on our project website, the source code, database, and outputs are detailed below. This project is still in progress, but is updated often.

Code structure

├── content (website)
    ├── overview.Rmd
    ├── methods.Rmd
    ├── analyses
        ├── hypothesis1.Rmd
        ├── hypothesis2.Rmd
        ├── hypothesis3.Rmd
├── data
    ├── dictionaries
        ├── preprocessing
            ├── compoundR.csv
            ├── polysemeR.csv
            ├── humanizeR.csv
        ├── h1_dictionary.csv
        ├── h2_dictionary.csv
        ├── h3_dictionary.csv
        ├── tree_data.csv
    ├── journal_rankings
    ├── regression_analyses
    ├── sensitivity_checks
    ├── text_results 
        ├── h1_results
        ├── h2_results
        ├── h3_results 
    ├── word_embeddings
├── src
    ├── 01_pubmed_db
        ├── 01_download_medline.sh
        ├── 02_pubmed_parser.ipynb
        ├── 03_clean_db.sql
        ├── 04_pubmed_abstract_db.sql
        ├── 05_filtered_publications.R
        ├── 06_articles_per_journal.sql
        ├── 07_articles_per_year.sql
        ├── 08_biomedical_abstracts.sql
        ├── 09_check_abstracts_tbl.sql
    ├── 02_text_trends
        ├── 01_hypothesis1.R
        ├── 02_hypothesis2.R
        ├── 03_hypothesis3.R
        ├── 04_all_hypotheses.slurm
        ├── 05_pub_figures.Rmd
        ├── supplementary_analyses
            ├── 06_aggregate_ids.R
            ├── 07_diversity_abstracts.sql
            ├── 08_diversity_abstracts.R
            ├── 09_soc_diversity_eda.R
            ├── 10_human_abstracts.R
    ├── 03_word_embeddings
        ├── 01_w2v_train.ipynb
        ├── 02_w2v_results.ipynb
    ├── 04_text_relations
        ├── unfinished_analyses
    ├── 05_collaborations
        ├── unfinished_analyses

Database structure

├── pubmed_2021
    ├── abstract_data
    ├── articles_per_journal
    ├── articles_per_year
    ├── biomedical_abstracts 
    ├── filtered_publications 

About

In this project, my colleague Catherine Lee (Rutgers) and I employ computational text analysis to examine quantitative trends in the use of diversity terms, OMB/Census terms, and other population labels in a sample of 2.6+ million biomedical abstracts spanning the last 30 years.

Topics

Resources

License

Stars

Watchers

Forks