Incipient reverse causality checks on life satisfaction vs. the financial one

In this paper, the author explored and analyzed the most relevant influences of financial and life satisfaction. They resulted from World Values Survey data between 1994 and 2020 (300K+ observations). Both Data Mining and Statistical methods supported this study. The first method used in this paper was the Adaptive Boosting in R via the Rattle interface. It benefited from the additional use of the LASSO package (Stata) to further focus on the most significant variables by performing random cross-validations and removing overfitting. Regressions such as binary and ordinal logit, probit, scobit, and Ordinary Least Square (OLS) served for further selections of the most robust predictors. Such choices stood on post estimations of collinearity, comparative accuracy, explaining power, and maximum probability via risk prediction nomograms. Some cross-validations on age, gender, and marital status additionally validated the final results. All led to the same conclusion based on comparative models in which these two types of satisfaction had both roles: input and response, as indicated by all tests. As expected, financial satisfaction proved to be a cause of life satisfaction rather than vice versa