Arrgh I hate studies like this.

[marydell]

Mothers' talk is key to kids' social skills, study says

1. we observed a particular behavior, let's call it behavior A, in some mothers
2. we observed a different behavior, let's call it behavior B, in the children of those mothers
3. therefore, Mothers engaging in behavior A will cause their children to engage in behavior B!

ARGH. No.  "Mothers with social skills produce offspring with social skills" might be a fair conclusion, or even "Mothers who talk about mental states produce children with social skills."  But saying that the talking is the means of producing the skills in the offspring is a leap. There could be another underlying cause for both the mothers' tendency to talk about mental states and the kids' tendency to have good social skills...genetics, for instance, or level of education, or whether they eat peanut butter, or whether they have other family members living at home.

(Here's my own theory, based on the same logic:  Mothers who wear boots in December are more likely to have children who wore coats in December than mothers who wear sandals in December. Therefore boot-wearing must cause coat-wearing! If we get the sandal-wearing mothers to wear boots instead, their children will start to wear coats!  Where should I go to apply for a grant?)

If they did a study in which they altered the way the mothers talk to their children and recorded observable results, with a control group etc, that would be persuasive.  This is just annoying and creates another way of blaming mothers for stuff, which is, of course, the thought that the artcle leads with. 

Comments

[jerusha.livejournal.com]

Shorter you: Correlation does not equal causation.

Also? Agree wholeheartedly!

[fledgist.livejournal.com]

Hear! Hear!

[seawasp]

I think,

"CORRELATION DOES NOT IMPLY CAUSATION"

needs to be tatooed on the forehead of every science reporter, and every scientist who ever TALKS to a science reporter.

Also, I applaud the idea of controlled human experimentation as a way of verifying guesses as to human behavioral causes and effects. For some reason it's outlawed, though.

[slrose.livejournal.com]

Not on the forehead of the science reporters -- on the back of their hand, where they might see it. If they're writers, that is.

If they're television reporters, the forehead will work fine.

[ashnistrike.livejournal.com]

Gah! I'm glad you get it, even if many of my students (and the reporter) don't.

My guess, based on the information as given in this article: Mothers with good social skills do a lot of things differently, including talking about mental states while looking at pictures. They probably also talk about their own feelings more, ask more about the kid's feelings, and engage in more sensible communication strategies when arguing with co-parents. All these things combine to model good theory of mind to the kid, improving their social skills in turn. The effect decreases as children are exposed to additional models. This is low on the shocking-findings scale.

Trying to train parental social skills is not a bad next step. Nattering on about blaming moms, on the other hand, not so much.

[marydell.livejournal.com]

Is it a given that good theory of mind is what gives kids good social skills? That seems like it leaves out, for example, people whose social skills are affected by neurological factors. What if everyone's social skills are a product of their neurological state?

(I realize I'm tossing half-baked ponderings at an expert...sorry, I can't help myself!)

[ashnistrike.livejournal.com]

Theory of mind is one factor out of several contributing to social skills, but it's an important one. If you can't run simulations of other people's mental states on your own OS, social skills become a matter of memorizing responses to particular behaviors, which is a PITA.

[rant] All mental abilities are affected by neurological factors; it's a different level of analysis. You can talk about skills and abilities in terms of the cognitive constructs that make them up, or you can talk about the brain areas--in the same way that you can describe what a painting portrays, or you can do a chemical analysis of the paint. Either may be useful depending on what you're trying to do, but they're not in conflict. [/rant]

Current theory likes mirror neurons as a neurological basis for theory of mind, but it's probably more complex than that.

[marydell.livejournal.com]

Hrm, interesting. I'm thinking of conditions like sensory integration disorder or autism, that cause people to miss social cues and therefore manifest poor social skills. Although that's probably the same issue again of using "social skills" to mean a variety of things. And now that I think about it, in those situations, social skills can be taught, it just requires different methods of teaching.

What's really needed is a study of identical twins, in which one twin from each set is raised by parents with good social skills, and the other is raised by parents with poor social skills, and then they're compared as adults! Darn those ethics. (I remember my archaeology prof joking about wanting to do a study raising one twin without any language, to test whether language is a distinct advantage in learning to make stone tools)

[ashnistrike.livejournal.com]

I was thinking of autism, too--the current theory is that it involves, among other things, deficits in the mirror neurons. Although the current theory is very excited about the mirror neurons in general, and we really need a few years' more research (and waiting for them to be a less sexy topic) in order to have any degree of certainty.

I heart your archaeology prof; we can always use more mad social scientists.

[ashnistrike.livejournal.com]

Another note: the problem is the reporter, not the original study. If I weren't in the middle of intensive final grading, I would look up the original peer-reviewed article--and expect that the claims they make there are not as extreme as those put forward by CNN.

[kyuuketsukirui.livejournal.com]

Yeah, this is almost always the case. Sometimes the original study is completely different from what is reported.

[fledgist.livejournal.com]

For "sometimes" read "frequently".

[elettaria.livejournal.com]

The impression I'm getting is "mostly". The ability of science reporters to turn around what was in the original research, sometimes 180 degrees, never ceases to astonish me.

[marydell.livejournal.com]

The study page is over here (http://www.esrcsocietytoday.ac.uk/ESRCInfoCentre/SearchResultsPage.aspx?data=z8HSvl3fWwX%2babs28k0zRHLQJ4hMqifc%2b7E5GVhPJeyEDjzvkqK5viJ65P%2bMuhzDtXThirRiqgDQSeOJ2ohbNU5p4xBunrVy6Fh2%2fvNVWYrzT1jt%2fOrEDSaiU4cV%2f%2bd1cB%2fOYl%2bTs9GLeaJJ0Eu1CYBNIWxXpoWrSI1JXWnW6ft8YkIm%2fuFVPnxkRk4wa8Vn5rZe9ZTUxzYySPjOGjj1BYaQxVr97r8iJa8uUaIiD5kUaCPibG1dYiIC3wDkamhp&xu=0&isAwardHolder=&isProfiled=&AwardHolderID=&Sector=&br=tr). I read the non-technical summary and it makes better sense, because it's specifically measuring children's understanding of State of Mind, rather than the nebulous "Social skills" used by laypersons.

However, it also says that the study is designed specifically to see whether SOM talk from mothers and SOM understanding in children is caused by the mothers, or by children with good SOM understanding eliciting better conversation from their mothers. The summary doesn't say anything about controlling for other possible factors, particularly genetic inheritance.

[ashnistrike.livejournal.com]

The full report doesn't say anything about genetic confounds either--I would guess that their sample is mostly genetically related, and that someone else would need to examine that factor.

At some point when I'm not procrastinating on grading, and therefore have more time, I should look at a couple of their papers and get more details on the sample and methods.

This sort of nuance almost never gets into science reporting, unfortunately. We have a science journalism major at IIT that is supposed to help with that--they started the same time I did, so we'll see how it goes in a couple of years, I suppose.

[browngirl.livejournal.com]

Ah, adventures in experimental design and science reporting. *applauds your analysis*