Abstract

Surveys of religious congregations are a mainstay of sociological research on organized religion in the United States. How accurate, reliable, and comparable are the data generated from the disparate methods used by researchers? We analyze four congregational surveys to show how two components of data collection—sampling design and survey response rate—may contribute to differences in population estimates between the surveys. Results show that in three populations of congregations (all religious traditions, Catholic parishes, and Hispanic Catholic parishes), estimates of key congregational measures, such as head clergy characteristics, congregational size, and Hispanic composition, are susceptible to differences in data collection methods. While differences in sampling design contribute to some of the variation in variable estimates, our unique analysis of survey metadata shows the importance of high response rates for producing accurate estimates for many variables. We conclude with suggestions for improving congregational data collection methods and efforts to compare survey estimates.

中文翻译：

---

宗教会众社会学研究中的调查数据收集方法和差异

摘要

宗教团体的调查是美国有组织宗教的社会学研究的主要内容。通过研究人员使用的不同方法生成的数据有多准确，可靠和可比？我们分析了四个会众调查，以显示数据收集的两个组成部分（抽样设计和调查回复率）如何导致调查之间人口估计的差异。结果表明，在三个会众（所有宗教传统，天主教堂和西班牙裔天主教堂）中，主要会众措施（例如，神职人员特征，会众规模和西班牙裔组成）的估计值容易受到数据收集方法差异的影响。尽管抽样设计的差异会导致变量估算值的某些变化，我们对调查元数据的独特分析表明，高响应率对于生成许多变量的准确估计值非常重要。最后，我们提出了一些建议，以改进公共数据收集方法并努力比较调查估计值。