I’m reading Froma Walsh’s Spiritual Resources in Family Therapy (1st edition) for my Wellness & Spirituality Throughout the Life Cycle class. Here’s a quote:

“Active congregational participation as well as prayer tend to become increasingly important over adulthood. Whereas only 35% of young adults aged 18-29 attend their place of worship weekly, 41% of persons aged 30-49, 46% of those  aged 50-64, and 56% of those over 65 attend weekly.”

That quote is from the 1999 edition of the book, and so those numbers are probably based on a survey conducted in the 1990s. The source is not cited, so I can’t be sure, so take this criticism with a grain of salt. I’m just using this example to point out something that happens a lot with the analysis of age-based research. That is, this presentation makes it sound as if humans attend church more and more as they get older, but these numbers say no such thing.

What these numbers say is that at the time of the survey, 35% of young adults say they are going to their place of worship weekly and that each age group above them at this time show more of that behavior. Each generation has its own characteristics. It may well be that this group of young adults is part of a less church-going generational cohort, which will stay more or less that way as they age. Imagine, for example, that such is the case and the next generation that comes along attends church more often. A survey at that time will show that place-of-worship attendance is relatively high in young adults, drops off in middle age, and then resurges in old age, and many will assume, based on that, that this is the “natural” progression of human church-going behavior.

As far as I know, Walsh is accurate in her analysis, based on information she is not giving. It’s a potential error to be aware of, though, and one often overlooked by researchers in psychology. I’ve noticed it often since reading Strauss & Howe’s Generations. They make the point really well, that we often think that increasing age causes people to become more or less something-or-other–more conservative, say–basing our reasoning on the generational cohorts that are currently alive, but it may just seem that way because of the quirks of our sample.

In order to know, we would need more information than this snapshot. We need multiple surveys conducted over quite a period of time, while different generational cohorts were alive, to get longitudinal information. Does each generation attend church more and more as it ages? Is the difference in church-going between a generation in its young years and that generation in old age greater or less than the difference between that generation and another generation entirely?