Read on Twitter
New research.
If you are a quantitative social scientist who ever uses a category for non-religion in your analyses (which is most of us), then this chapter will hopefully be useful. https://global.oup.com/academic/product/empty-churches-9780197529317#.YA8vtlCxGi4.twitter
If you are a quantitative social scientist who ever uses a category for non-religion in your analyses (which is most of us), then this chapter will hopefully be useful. https://global.oup.com/academic/product/empty-churches-9780197529317#.YA8vtlCxGi4.twitter
The basic idea here was a RELTRAD for secular people, using inductive analyses to identify the best way to classify secular survey respondents. I used the 2014 Pew Religious Landscape Survey because it is large enough to have decent sizes for small categories /2
Although how many categories you can create for secular respondents is clearly dependent upon sample size and available measures, I lay out some rules of thumb for classification that can be helpful in most cases. /3
Rule 1: Most importantly, always separate atheists and agnostics from non-affiliated theists. These groups are very, very different along a number of sociological dimensions. /4
Rule 2: If you have enough cases, separates atheists, agnostics, and nonaffiliated theists into categories. Cluster analysis supports this tripartite classification scheme.
(Whew! Would have been a bummer to publish a paper showing my previous book was garbage...)
/5
(Whew! Would have been a bummer to publish a paper showing my previous book was garbage...)
/5
Rule 2a: People who say “nothing” to an identity question and “no” to a question about god should be classified with people who openly identify as atheists. /6
Rule 2a.2: BUT, keep these “reticent atheists” and open atheists separate if you can too, because there are distinct sociological patterns to these different approaches to non-theism, as I outline here: https://www.degruyter.com/view/book/9781479838271/10.18574/9781479838271-015.xml
/7
/7
Rule 3: Examine religiously affiliated non-theists as a possible category. Overall, these individuals tend to look more like other atheists than they do other religious affiliates. Combining with other atheists is viable, depending on your research question. /8
If you are designing a survey, for the love of (no) god, make sure to ask questions about both secular identification and belief in god. Provide at least an atheistic and agnostic answer option to the god question. /9
Overall, we need to be more rigorous about how we treat secular respondents to surveys, and avoid lumping people into a single category if at all possible. Preprint available here: https://www.researchgate.net/publication/348755419_The_Many_Meanings_of_the_Secular
DM or email me for the legit PDF.
/10
DM or email me for the legit PDF.
/10
By far the best thing about this contributing to this volume was the chance to work with such smart and constructive scholars, including @ntammerman, @HedstromMatt, @drmacgregor, and many others (wisely) not in the Twitterverse. /11
Best of all was the chance to spend time with the late, wonderful Vern Bengtson. A more accomplished scholar and kinder soul you could not find. A model for us all. RIP Vern, thanks for being an exemplar.
