Human Wellbeing and Machine Learning

Supplementary Material for

Human Wellbeing and Machine Learning

by Ekaterina Oparina (r) Caspar Kaiser (r) Niccolò Gentile; Alexandre Tkatchenko, Andrew E. Clark, Jan-Emmanuel De Neve and Conchita D'Ambrosio

This repository contains the list of variables that are used in the Extended Set analysis for the German Socio-Economic Panel, the UK Household Longitudinal Study, and the American Gallup Daily Poll. The variables are grouped into categories, the summary table is reported at the beginning of the document. We use the 2013 Wave of Gallup and SOEP, and Wave 3 of the UKHLS (which covers 2011-2012). Our dataset includes all of the available variables, apart from direct measures of subjective wellbeing (such as domain satisfaction, happiness, or subjective health) or mental health and technical variables (e.g. id numbers). We also exclude variables with more than 50% missing values.

The presented lists include the variables before processing. For the analysis, we convert categorical variables into a set of dummies, one for each category. We then drop all perfectly collinear variables.

《人类福祉与机器学习》补充材料

作者：叶卡捷琳娜·奥帕里娜（Ekaterina Oparina）(r)、卡斯帕·凯撒（Caspar Kaiser）(r)、尼科洛·真蒂莱（Niccolò Gentile）；亚历山大·特卡琴科（Alexandre Tkatchenko）、安德鲁·E·克拉克（Andrew E. Clark）、扬-埃马纽埃尔·德内夫（Jan-Emmanuel De Neve）与孔奇塔·丹布罗西奥（Conchita D'Ambrosio）

本仓库包含用于德国社会经济面板（German Socio-Economic Panel, SOEP）、英国家庭纵向研究（UK Household Longitudinal Study, UKHLS）以及美国盖洛普每日民意调查（American Gallup Daily Poll）扩展集分析的变量清单。所有变量已按类别分组，文档开篇即附变量汇总表。

本研究采用2013年版盖洛普每日民意调查与SOEP数据集，以及覆盖2011-2012年的UKHLS第3波数据。本数据集纳入全部可用变量，但排除以下几类：主观福祉直接测度指标（如领域满意度、幸福感或主观健康状况）、心理健康相关指标以及技术类变量（例如身份编号）；同时我们还剔除了缺失值占比超过50%的变量。

此处展示的变量清单为处理前的原始版本。在分析流程中，我们将分类变量转换为一组虚拟变量（每个类别对应一个虚拟变量），随后剔除所有完全共线性的变量。