Mapping the Statistical Significance of Factors Contributing to the World Happiness Report

Abstract

This paper aims to delineate findings of the statistical significance of the factors contributing to the happiness score. The happiness score also termed as ladder score, is a metric used by the United Nations Sustainable Development Solutions Network to metricize the happiness of the citizens in a country. To tackle this issue, we use regression and data visualization. We perform a survey on the factors affecting ladder score and how these factors can be used for predictive analytics. We use Linear Regression, Polynomial Regression, Lasso Regression with cross-validation, and Ridge Regression with cross-validation. Next, we use evaluation metrics like MSE, RMSE, Adjusted r-squared, and r-squared value for the evaluation of the factors on the predictive model. Then, we plot the countries mentioned in the report on a geographical scale based on their happiness index scores. Furthermore, we plot the statistical significance of these factors on a continental scale, to reveal insightful patterns over a larger geographical domain. We aim to bring to light the trends of the aforementioned factors and produce the significance of these results on a world map. The results will help elucidate the global patterns formed by these metrics. An additional application is an extrapolation of the results procured. To augment the metrics of the Word Happiness Report in a statistically comprehensive way. Furthermore, through this evaluation, the world happiness report can be revised to accommodate more inclusive factors and mitigate the redundancy of the factors.