This repository contains all the main scripts and data that are needed to reproduce the analysis that we report for the iconicity ratings in the following paper:
Winter, Bodo, Gary Lupyan, Lynn K. Perry, Mark Dingemanse, and Marcus Perlman. “Iconicity Ratings for 14,000+ English Words.” Behavior Research Methods, April 20, 2023. https://doi.org/10.3758/s13428-023-02112-6.
For all analyses using our iconicity ratings, we recommend using the iconicity_ratings_cleaned.csv file in the ratings folder. This contains the files that are processed.
The main script folder contains the following files:
-
processing_raw_ratings.Rmd: takes thecombined_byWord_and_subject_anonymized.csvfile, applies exclusion criteria and aggregates the data to yield the cleaned file, which is the main iconicity ratings file that will be used for all analysesiconicity_ratings.csv -
bayesian_models.Rmd: takesiconicity_ratings.csvand performs a series of analysis, specifically correlations with other lexical measures from published studies that are in theadditional_datafolder -
bayesian_models.Rmd: posterior predictive checks for all models that will be saved in the folderpp_checks -
substantive_analysis.Rmd: the main analysis script that reproduces all the main results reported and creates plots (this script loads in the models created bybayesian_models.Rmd) -
calculate_ICCs.Rmd: inter-rater reliability measures added after review