Also, all these numbers are just painfully tidy.

This changes changes their “results” from “so unlikely they are pretty much impossible” to a straight “impossible”.

If there is no subset with a high happyness with marriage ratio of over 26% and the merger of those subsets is the whole population (of wives in this case). The high happyness with marriage ratio CAN NOT BE over 26%.

I guess you have heard of gerrymandering. You might have asked yourself why he didn’t tailor the districts in a way that ALL of them are in his hand. Well, because it’s impossible when a majority of voters are against him.

Same thing here. If the whole population of wifes has a high happyness with marriage ratio of at least 50%, then at least one of the subsets by number of children has to have 50% too, as every wife has a number of children and therefore is in one of the subsets.

So this has gone from “Ok, I’m almost certain this is fake.” to “It is absolutely impossible for this not to be faked numbers.”

Those are pretty generous for that christian “scientists” to say the least.

To make our assumptions even more generous: Even if there are no married couples with 4 children so all couples would have either 0 or at least 5 children*, the average number of children per married woman would still have to be 2.6, which is quite a bit higher than the 2.0 or so children per woman OECD gives for the US.

*Yes, this would imply that woman always get at least 5 children AT ONCE at first birth giving and no children would die before “replacement” was born. Didn’t i say these were rather “unlikely” assumptions to make the best case for them?

Look at the average high divorce proneness in husbands in the whole population from the second graph. Obviously you can’t tell what it is excatly because you don’t know how many people feel whether god is the center of their marriage. But you know that it is at most 11%.

Now from the first graph we know that 21% of the husbands in marriages in which not both partners attend church regularly have a high divorce proneness and it’s 6% for husbands in marriages where both partners regularly attend church.

We can now calculate how many married couples have to both attend church regularly so that the average high divorce proneness in husbands is at most 11%.

Let x be the ratio of married couples who both attend church regularly per married couples, then:

6% * x + 21% *(1-x) x > 2/3

So at least two thirds of married couples attend church together* regularly.
If that was true churches wouldn’t have to lie so obviously…

Now … let’s come to the third graph …
to put it simply:

For this and the other two graphs to be true at the same time a LOT of married couples would have to have at least 5 children. And by a lot i mean AT LEAST HALF OF THEM. This is obviously not true in the american society as a whole and it’s nigh imposible to draw such a subset of the americans in a representative way. (minor concern: why don’t you list 5+ children if that is more than half your sample?)

TL:DR results are mathematically about as unlikely as you quantum tunneling to the other side of the planet in the next five minutes.

*They don’t have to go together, but both partners in a marriage attending church regularly will result in them doing so together rather often, i guess.

High divorce proneness was defined as
reporting a “3” or higher (on a scale from 0-10) regarding
the likelihood that “you and your partner will eventually
separate or divorce.”

So it’s (a) self-determined, and (b) set at 30% likeliness to be “highly prone”.  Thanks, science!

Good post!

It’s ONE POSSIBLE explanation, but it could be that just doing things together–i.e., having common interests–makes the marriage happier. If it was religious attendance, specifically, I’d expect there to be a slight bump in the numbers when only one spouse attends regularly.