How should we measure people’s perceptions of—and attitudes about—economic inequality? A recent literature seeks to quantify the level of inequality that people, especially Americans, perceive and prefer in society. These findings have garnered much attention from both social scientists and the public. But many of the methods used in this literature are either known to have methodological issues or have not been thoroughly compared against other methods. Thus it is not clear which, if any, are valid and reliable measures of perceived, or preferred, inequality. To assess these measures, we conducted a large web-based study (N = 831) to compare key methods for measuring perceived inequality and their related justice attitudes. In addition to comparing the resultant summary statistics, we assess how well the different measures correlate with each other and with Likert scale measures of perceived inequality. Our analysis reveals a range of issues with these measures, including failure to provide logical responses, large method effects on point estimates of inequality, and low correlations between methods and with criteria measures. We conclude our analysis with three recommendations for researchers aiming to measure inequality perceptions and preferences.

我们应该如何衡量人们对经济不平等的看法和态度？最近的文献试图量化人们（尤其是美国人）在社会中所感知和偏爱的不平等程度。这些发现引起了社会科学家和公众的极大关注。但是，已知该文献中使用的许多方法都存在方法学问题，或者尚未与其他方法进行彻底比较。因此，尚不清楚哪一种（如果有的话）是感知到的或偏好的不平等的有效和可靠的度量。为了评估这些措施，我们进行了一项基于网络的大型研究（N = 831），以比较衡量感知的不平等及其相关司法态度的关键方法。除了比较结果汇总统计量之外，我们还评估了不同度量之间的相关程度以及与感知不平等的李克特量表之间的相关程度。我们的分析揭示了这些措施的一系列问题，包括未能提供逻辑响应，对不等式的点估计产生较大的方法影响，以及方法与标准措施之间的相关性较低。我们以针对研究人员的三项建议作为分析的总结，这些建议旨在衡量不平等观念和偏好。