The big contribution of this book is a survey, the Next Mormons Survey (NMS), of current and former Mormons. Riess has done a major public service by conducting this survey and providing some resulting tabulations. A key aspect of the survey is that its designers set quotas of how many current Mormons and how many former Mormons they have in the data (though they fell far short of their goal for former Mormons). That is going to matter for some of the inference, as I'll note below. But she compares her results to other survey results when possible (GSS, Pew, etc), and they seem to hold up reasonably well, at least among the ones included in the book.
My main concern with the book is that I'm very unsure about a lot of the statistical inference. It is really unclear how seriously we should interpret results of narrow tabulated cells, where sample sizes become very small (as Riess acknowledges). But even aside from that, I have concern about the inference. An example: on page 106, Riess writes, "The NMS confirmed that Mormons with college degrees are 15 percent less likely to leave the church." If you paid attention above, you'll know that this claim cannot be made with the NMS, which targeted its current/former Mormon split. What you *can* observe in NMS is the probability that a current Mormon has a college degree, and the probability that a former Mormon has a college degree. I strongly suspect that in Riess' sample, current Mormons are 15 percent (or maybe percentage points? I'm concerned) more likely to have a college degree than are former Mormons. That does not mean that a Mormon with a college degree is 15 percent less likely to become a former Mormon. You could get to that number with Bayes' Rule, but the NMS doesn't give us all the info we'd need for that.
There are lots of examples like this where I suspect Riess is getting the inference wrong (I emailed her asking about this; I will revise this review if she responds). Or, if she's getting it right, she's using more outside information than it appears.
There are other issues too, such as places where she draws inference without giving sufficient attention to selection effects. And there are places in the book where she uses the term "Mormon" ambiguously, such that I'm not sure if the data she's referring to are for the whole sample or just the current Mormons.
So, I have concerns about the inference and that makes me nervous about how much I should learn from the book. That said, I very much enjoyed reading it, and it presents some interesting results. Some results from the book that I found interesting:
- Mormons living in Utah are more likely than Mormons outside Utah to embrace strong faith statements, report that they are active in church, hold current Temple Recommend, and so on.
- Half of millennial Mormons embrace young earth theory and reject evolution! This surprised me. The share is 60% for baby boomers.
- A bit over half of millennials are bothered by the male-only priesthood (versus a quarter of boomers).
- About 70 percent of Mormons, of all generations, agree with the November 2015 policy defining homosexual relationships as apostasy. About 60 percent of all generations agree with the ban on children of same sex couples being baptized. These may seem high, but 1/3 of members opposing that controversial policy seems notable.
- Word of Wisdom compliance is lower than I thought. 40% of self-reported "active" Mormons break it in some way.
- Mormons' acceptance of homosexuality is rapidly rising, due largely to millennials, more than half of whom don't have a problem with it.
So, take the above with a grain of salt, but some very suggestive results.
The most annoying thing about the book is that the endnotes are organized by chapter number, but within the text, there is no heading telling you which chapter number you are in. You have to flip back to the beginning of the chapter to figure it out, before you can go look up the end note. That's bad! Don't do that, publishers!
Overall, a pretty good read--very engaging, it really was a page turner. She includes a lot of anecdotes from in-depth interviews she conducted (separately from the NMS), which I really enjoyed. I would have liked much more care to be given to discussing the stats precisely.