LogFAQs > #970074757

LinkPizza posted...

To me, it is (at least in this case)...

That means you are grossly overestimating your own competence.

LinkPizza posted...

Especially when I don't even know how they came to believe which factors are causing what they see in their data...

If you actually read into and understand studies, competently-written ones for subjects like this usually outline different variables that they considered and break down their observations into various categories, such as race, sex, socioeconomic status, and region.

The biggest and easiest one to look for, though, is sample size: If you sample 10 people, 5 of whom were spanked and 5 of whom were not, finding that 3 of the spanked people developed problems and only 1 of the control group did not isn't an overly meaningful correlation, since on that scale random variation in other factors can indeed create correlations without there being any sort of causal relationship. If you sample 1,000,000 people, though, and 300,000/500,000 of the spanked group displayed problems and 100,000/500,000 of the control group did not, that's a very solid basis for saying that being spanked triples one's risk of developing those negative effects unless you can find another factor that's similarly correlated.

Meanwhile, self-reported anecdotal evidence from those who were spanked or have spanked their own children is largely useless. Those making the assessments are doing so from a layperson's perspective (meaning they can't necessarily recognize the potential negative effects), and the assessments carry an extreme risk of bias given that most people don't want to think that they or their parents might be child abusers.

The information you are observing is objectively without merit. Basing opinions on it is a terrible idea by every possible metric. So stop it.