[compiler-guy]

That's the thing about trends in aggregate data. It tells very little about the details of any particular situation. There are almost certainly a wide variety of subgroups where this particular trend doesn't hold, and others where the changes are even more dramatic.

But the aggregate trend is quite clear.

[IcyWindows]

Isn't the point that there is no way to get reliable data on this?

How would aggregation of unreliable data help?

[compiler-guy]

The data is reasonably reliable, at least in the view of the team that published the paper, and those who reviewed it. No one claims it is perfect.

And the post that started this sub thread was about how their experience didn’t show the trend. But no one in the social sciences expects every sample to follow the trend. There will be numerous exceptions. Just like sometimes when one rolls a pair of dice one gets a twelve.

That twelve is an absolutely an accurate sample from the data but just because one sometimes gets an outlier it doesn’t mean that there is no central tendency.