Language experience predicts music processing in a half-million speakers of fifty-four languages

This is the repository for Liu, Hilton, Bergelson, & Mehr (2023, Current Biology). The paper is available at https://www.cell.com/current-biology/fulltext/S0960-9822(23)00387-1 and the preprint is at https://www.biorxiv.org/content/10.1101/2021.10.18.464888v2.

This repo contains:

• an R Markdown file that generates the manuscript
• analysis, and visualization code to produce the results reported in the manuscript
• a script that automatically downloads the data required to run the analyses and visualizations

Further data and information are available elsewhere:

• the data required for the analyses and visualizations is at https://zenodo.org/record/7614189#.Y-LAp-xBz0p
• the pre-registration is available at https://osf.io/xurdb.
• you can participate in the music perception test at https://themusiclab.org/quizzes/miq.

For assistance, please contact the corresponding authors: Jingxuan Liu (jl6297@gsb.columbia.edu), Courtney Hilton (courtney.hilton@auckland.ac.nz), and Samuel Mehr (sam@yale.edu).

Anatomy of the repo

Upon first downloading this repository, you should run the data_downloader.R script in the home directory. This will download all the required data (from https://zenodo.org/record/7614189#.Y-LAp-xBz0p) and move it to the correct locations.

After you have done this, to render the paper, run the code in /writing/manuscript.Rmd.

Warning
The manuscript file combines output from several .Rmd files devoted to analysis, visualization, and the like. The full_run flag in manuscript.Rmd determines whether analyses and figures should be generated from scratch (which can take > 30 minutes), or not. By default, it is set to FALSE, to save knitting time. If you set it to TRUE, all preprocessing, analysis, and visualization code will be run.

Data and analysis code

/data contains all the data (the data_downloader.R needs to be run to fully populate):

• /Exploratory contains the exploratory datasets: exploratory pre-exclusion data (Explore_full.csv), exploratory post-exclusion data (Explore_filtered.csv), exploratory one-to-one matched data (Explore_matched.csv), and exploratory inverse-probability weighted data (ipw_explore.RData).

• /Confirmatory contains the confirmatory datasets: confirmatory pre-exclusion data (Confirm_full.csv), confirmatory post-exclusion data (Confirm_filtered.csv), confirmatory one-to-one matched data (Confirm_matched.csv), and confirmatory inverse-probability weighted data (ipw_confirm.RData)

• /Combined contains the combined dataset (Combined_filtered.csv), needed for the main analyes.

• Language features and classification (language.csv)

• Headphone check scores (headphone_scores.csv)

• /meta-analysis contains the data used in the meta-analysis.

/results contains all the pre-saved results from previous runs of the analysis scripts (the data_downloader.R needs to be run to fully populate):

• analyses.RData contains the results from the analysis/analysis.R script.
• meta-analyses.RData contains the results from the analysis/meta-analysis.Rmd script.
• permutation_tests.RData contains the results from the permuted Discriminant Function Analysis, also in the analysis/analysis.R script.

Visualizations

Visualization code is in /viz, along with images and static data used for non-dynamic visualizations. The /viz/figures subdirectory contains static images produced by figures.Rmd, which can be regenerated with a full_run of manuscript.Rmd.