Scholarlink

To demonstrate the framework and process of implementing the proposed Computing Grounded Theory (CGT), this chapter takes subjective well-being as an empirical example of Y to find the most powerful predictors that can discover its potential “new” explanatory variables. Using the 2017 China General Social Survey (CGSS2017), this study illustrates the framework and process of how to train machine learning models of Y (i.e., subjective well-being) to estimate the SHAP values of each predictor. Complementarily, data from the World Values Survey (WVS) are employed to examine the main determinants of subjective well-being worldwide and to compare cross-national variations. Grounded in these empirical results, five application values of CGT are summarized. The results show that the data-driven, algorithm-based framework of CGT demonstrates how supervised machine learning can advance quantitative social science research, facilitating theory generation directly from data.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Applications of Computing Grounded Theory: Revisiting Subjective Well-Being

  • 章节

  • 出版模式：

    开放获取

  • 发表日期：

    2026年

查看全文
  • Yunsong Chen,
  • Zhuo Chen,
  • Wen Ma,
  • Guodong Ju

摘要

To demonstrate the framework and process of implementing the proposed Computing Grounded Theory (CGT), this chapter takes subjective well-being as an empirical example of Y to find the most powerful predictors that can discover its potential “new” explanatory variables. Using the 2017 China General Social Survey (CGSS2017), this study illustrates the framework and process of how to train machine learning models of Y (i.e., subjective well-being) to estimate the SHAP values of each predictor. Complementarily, data from the World Values Survey (WVS) are employed to examine the main determinants of subjective well-being worldwide and to compare cross-national variations. Grounded in these empirical results, five application values of CGT are summarized. The results show that the data-driven, algorithm-based framework of CGT demonstrates how supervised machine learning can advance quantitative social science research, facilitating theory generation directly from data.