: Validity on a study about spanking children Earlier today, I got into an argument with someone about spanking children and whether or not it caused them to become more violent later in life in addition to other issues. Eventually,

Vote Up Vote Down   0 Reactions

Login to give Reaction!

TL;DR read the original paper and other literature and come to your own conclusion.

Generally speaking, you have to remember that correlation does not mean causality. It can be very difficult to tease out confounding factors in a properly controlled study, let alone social sciences papers where controls aren't really feasible.

Beyond that you shouldn't trust the article you linked to, go back to the original source which I think is this presentation. Two things really struck me reading through it.

First was how inconsistent some of the data is. For example, slide 5 says 33% of Swedish children are spanked daily... but slide 8 says the number of "frequently spanked" children in Sweden is 17%. I'm not necessarily surprised (I take a rather dim view of the rigour of published research in general to begin with), but this does speak to the challenges this kind of research faces.
Second was an absence of an attempt to find or acknowledge confounding factors. For example, whether or not factors associated with criminality like socio-economic status are also correlated with levels of spanking. This kind of analysis may be somewhere, but I haven't a hot clue where to find it.

All that being said, I can sit down and pick apart pretty much any paper out there, but my analysis will say as much about my world-view as anything else. Go to the guy's webpage (there is a link in the original article you cited) and read his papers with a critical eye. Then search the literature for papers on the benefits of spanking, and the negatives of alternatives to spanking, and come to your own conclusions.