This component contains all public data files used in this project. For readers who are interested in reproducing the analyses reported in the report's main text or supplement, please visit the "Analysis" component, which contains .zip packages with these same data sets and accompanying R scripts.
Note. We have taken several steps to protect the identities of respondents who participated in this study. These steps included:
(1) blurring all actual USCF and FIDE ratings that were obtained as part of this research to the nearest 25-point increment,
(2) removing all individual-level demographic information about sex, race/ethnicity, age, educational attainment, income, employment status, and whether a given player held a provisional rating. This means that some analyses, rerun on the publicly available datasets, will not produce the exact values reported in the manuscript and supplement. Some analyses (i.e., those run using demographic variables) will not run at all, but we include the analysis code used to generate the results we report for transparency. We have confirmed that analyses run on deidentified (i.e., blurred) data produce results similar in directionality, significance, and magnitude to results run on the original data. Information on which variables have full information, partial (blurred) information, or are removed from the public archive, is available in ListOfPublicVariables.pdf.
(3) removing the datasets used to match USCF and FIDE IDs to the USCF and FIDE database files when obtaining actual tournament ratings at the time of the survey and six/twelve months later.
Interested researchers can contact Patrick Heck at pheck1000@gmail.com with any questions about the public archive, analyses, or privacy-protected data not made public.