Abstract: In the past two decades, scholars have devoted much attention to the measure of Chinese religions, mainly using the scheme based on denominational affiliation, which is the most common approach to religious classification in western societies. However, the denomination-based scheme cannot capture the actual religious life of China. We point out four challenges this scheme encounters in survey research in China: the foreignness of the Chinese term ‘religion’ (Zongjiao); the misconception of denominational affiliation; the inapplicability of compulsory, one-single-choice religion; and the social or political sensitivity of specific religions, especially Protestantism. After critiquing the traditional scheme used to measure Chinese religions, we offer a new approach that addresses its shortcomings. Our revised approach attempts to research belief without using the term ‘religion’, focuses on belief in deities rather than on denominational affiliation, and allows multiple answers to the question about religious beliefs. In order to compare the denomination-based scheme with the deity-based scheme, we conducted experiments in the three waves of the China Family Panel Studies in 2012, 2014, and 2016. Our results show that the deity-based scheme yields more meaningful interpretations and more accuracy in religious classification than the denomination-based scheme in China. This article ends with some suggestions for improving the measurement of Chinese religion in future survey research studies.

Key words: Measures of Chinese religions, denomination-based scheme, deity-based scheme, China Family Panel Studies