Data Repository
Description
We have used the dataset of Joint EVS/WVS 2017-2020, allowing for comparisons of tax morality in more than 79 countries worldwide, applying machine learning variable selection methods suited to categorical variables (Chi-Squared and mutual information). The empirical results revealed that variables like: religious denomination, faith in God and the importance of God, together with the level of trust in people from other religion, confidence in churches have been connected with a significantly increased the extent of tax morale, being the most relevant variables of religiosity.
Files
Steps to reproduce
The data that support the findings of this study are openly available in GESIS DATA ARCHIVE at http://doi.org/[10.4232/1.13560]. EVS (2020): European Values Study 2017: Integrated Dataset (EVS 2017). GESIS Data Archive, Cologne. ZA7500 Data file Version 4.0.0, doi:10.4232/1.13560. We have used the dataset of Joint EVS/WVS 2017-2020, allowing for comparisons of tax morality in more than 79 countries worldwide, applying machine learning variable selection methods suited to categorical variables (Chi-Squared and mutual information) using R tool.