Interesting Links for 10-11-2014

[andrewducker]

• Sunshine can slow weight gain and diabetes symptoms - irrespective of Vitamin D levels (In mice)
  (tags: sun weight health vitaminD mice )
• 'MI6 chief and shamed diplomat raped me' claims alleged victim of VIP paedophile abuse ring
  (tags: mi6 abuse rape )
• Dear America, when your police are instituting no-fly zones to prevent media reporting of

Comments

[nancylebov]

Would you be willing to mark studies based on simulated situations the way you mark rodent studies?

I feel like I want to go out and lie to a scientist, a journalist, and a headline writer.

[andrewducker]

The situations aren't being simulated here. The children were actually lied to, and being lied to actually caused them to lie more later on.

[doubtingmichael]

It still seems like kind of a simulation here. It shows that if a stranger lies to a child, they are more likely to lie back to that stranger a few minutes later. I'm not saying it's a bad study, but going from its evidence to "parents, if you lie to your kids, they'll become liars" is quite a jump.

Clearly, it would be much better to find some pathological liars and ask them what their children are like. There is no possible failure mode with that idea.

[andrewducker]

I'm _totally_ open to "That's a hyperbolic headline." - and the article agrees that it's a stretch.

But pretty much all experiments are simplified, because it's the only way to isolate variables. So if people don't like that, I'd rather they just didn't read anything to do with any social science.

[doubtingmichael]

Well, I'm interested in what they are saying, I recognise that simplication is necessary, and I'm happy with a lot of social science. But sometimes I'm interested to poke critically at a study, and think about its limitations.

[andrewducker]

Me too - and I'm very happy for people to talk here about limitations of things I link to. I actively appreciate it!

[doubtingmichael]

Speaking of science, have you seen this?

http://www.bbc.co.uk/news/world-europe-29996872

Good news, I think.

[andrewducker]

Yes, reversing some massive stupidity there.

[nancylebov]

Here's what the article said:

""The actions of parents suggest that they do not believe that the lies they tell their children will impact the child's own honesty," the researchers said. "The current study casts doubt on that belief. This study suggests, rather, that [school age] children may use the actions of adults, as a model, to determine whether they will engage in honest or dishonest behaviours." They added: "Perhaps adults need to re-evaluate the way that they interact and talk with children.""

It's generalized beyond interactions with strangers.

I'm not sure how you'd find out what goes on between parents and children, though it might be interesting to ask parents and children about how much they lie to each other. Unfortunately, how good your information is will depend on whether your subjects trust you to keep the information confidential.

[andrewducker]

I agree - it is, hence me being happy about accusations of hyperbole.

The article also says:
"One limitation of the study, acknowledged by the researchers, is that the lying to the children was done by a stranger, not by their parents. It's possible that children might respond differently to parental lies - perhaps making it more or less likely that they will follow suit."