Chinese Journal of Sociology ›› 2019, Vol. 39 ›› Issue (3): 183-209.

Previous Articles     Next Articles

Explaining One Subjective Variable with Another: A Methodological Clarification

HU Anning   

  1. Department of Sociology, Fudan University
  • Online:2019-05-20 Published:2019-05-20
  • Supported by:

    This study was supported by the Fudan First-Class Research Group Funding (IDH 3458007) and the National Youth Talent Support Program, and the Shuguang Scholar Project of the Shanghai Education Committee (17SG08).

PDF

745

Abstract:

Sociological studies have rich subjective variables such as attitudes, perceptions, and preferences, and explanations of this type of variables often call for some more subjective variables. Empirical sociologists often have concerns about this approach of "subjective variable explaining subjective variable" in their empirical studies. Thus far, it is still not clear what kind of shortcomings the approach of "subjective variable explaining subjective variable" has, and more importantly, how to handle these concerns in order to enhance empirical explanations. Against this background, this study investigates the potential confounding bias in the "subjective variable explaining subjective variable" approach, and presents three strategies of theoretical reasoning, variable measurement, and data analysis in dealing with the issues. For each strategy, this article provides examples and discusses its advantages and limitations. Some unsolved issues, such as the direction of causation, the sensitivity analysis, and the measurement error, are also explored here. This study intends to bring in more diversity of sociological approaches and enhance the interdis-ciplinary collaboration (e.g., with psychology and political sciences).

Key words: psychological mechanisms, confounding bias, fixed effects, subjective variables