While there has been a great deal of media focus recently on the rise of those without religious affiliation (also known as the “nones”), there is an underlying issue facing this line of research: different surveys come to completely different conclusions about how many nones actually exist in the United States. Using the General Social Survey (GSS) and the Cooperative Congressional Election Study (CCES) this work details how each of these instruments measures religious affiliation in a different manner and how that results in an estimate of the nones that diverges by over 8% points in 2018. Statistical analysis reveals that the GSS has a much higher share of Protestants who never attend church than that found in the CCES. In addition, the CCES Protestant subsample is more Republican, while the nones in the GSS are more to the left of the political spectrum than the nones in the CCES. Some advice and caution is offered to researchers who are interested in studying the religiously unaffiliated in these two surveys.

中文翻译：

---

有多少“无”？解释调查估计的差异

虽然最近有大量媒体关注无宗教信仰者（也称为“无宗教信仰者”）的崛起，但这一研究领域面临一个潜在问题：不同的调查得出了完全不同的结论，即有多少在美国实际上不存在。使用一般社会调查 (GSS) 和合作国会选举研究 (CCES)，这项工作详细说明了这些工具中的每一种如何以不同的方式衡量宗教信仰，以及这如何导致对差异超过 8% 的估计2018. 统计分析显示，与 CCES 相比，GSS 中从未上过教堂的新教徒比例要高得多。此外，CCES 新教子样本更像是共和党人，而GSS中的nones比CCES中的nones更偏向于政治光谱的左翼。在这两项调查中，向有兴趣研究无宗教信仰的研究人员提供了一些建议和警告。