Cross-national analyses require additional controls to account for the non-independence of nations

This repository contains the data and R code for the manuscript "Cross-national analyses require additional controls to account for the non-independence of nations".

Getting Started

Cloning this repository

To clone this repository using Git, you will need to first install Git LFS to download large data files. Once Git LFS is installed, run:

git clone https://github.com/ScottClaessens/crossNationalCorrelations

Installing required packages

To run this code, you will need to install R and the following R packages:

install.packages(c("brms", "cowplot", "conleyreg", "countrycode", 
                   "dagitty", "geosphere", "ggcorrplot", "ggdag", "ggrepel", 
                   "ggtext", "haven", "lmtest", "papaja", "psych", "readxl", 
                   "rstan", "sf", "sjlabelled", "targets", "tarchetypes",
                   "tidybayes", "tidyverse"))

If you run into issues with the pipeline, try installing the specific R version or package versions outlined in sessionInfo.txt.

Executing code

1. Set the working directory to this code repository
2. Load the targets package with library(targets)
3. To run all analyses, run tar_make()
4. To load individual targets into your environment, run tar_load(targetName)

Accessing geographic and linguistic distance matrices

The geographic and linguistic distance matrices used in the paper can be found in the data/networks folder of this repository (see here). The two .xlsx files are as follows:

• 1F Population Distance.xlsx contains geodesic distances between 269 countries, measured in metres.
• 2F Country Distance 1pml adj.xlsx contains linguistic distances between 269 countries, scaled between 0 and 1. Higher values indicate greater linguistic dissimilarity between countries.

Help

Any issues, please email scott.claessens@gmail.com.

Authors

Scott Claessens, scott.claessens@gmail.com