The Race Test

[Placid Dingo]

The kin selection thing isn't consistent with what the link says;

black children also have a bias toward white but not nearly as strong as the bias shown by the white children.

I started doing an article on a Jewish Neo-Nazi (later changed my subject) for high school, and came across other research looking at how attitudes towards race can be easily adopted by kids viewing the same kinds of steriotypes. They talked about black kids treating golliwog dolls worse than the white dolls as one example.

[Requia ☣]

While the CNN link says that, it doesn't seem to appear in the actual study.

Or not!  They released an expanded paper, with all the questions and most of the data.  Useful, as the first paper failed to explain key terms in the results section.  I can see that in the paper now (though, it only seems to apply to some of the questions, black kids apparently want black kids as classmates, but not as friends* for example).

Also, I now have a better idea of what exactly 'self selected skin tone' is, its the skintone the kid wants to have.  There's a bias at play here, the things you want always seem better than the things you don't want (it occurs to me I know next to nothing about the ontology of biases, this may not be present in children).  It seems a strange way to manipulate the data too.

What really bugs me now is that half the questions are apparently about what adults or teachers think.  Questions about what a person believes and what a person thinks other people believe get different answers.  People are both more able to recognize bias in others, and have a bias against admitting to traits seen as negative.  Its possible this doesn't apply to children, but this is still sloppy.

Also, to address an earlier point of Ratatosk's, a number of the kids actually did refuse to answer the questions, sometimes as many as half the kids for a given question.

*Hyperbole.

[Bebek Sincap Ratatosk]

While the CNN link says that, it doesn't seem to appear in the actual study.

Or not!  They released an expanded paper, with all the questions and most of the data.  Useful, as the first paper failed to explain key terms in the results section.  I can see that in the paper now (though, it only seems to apply to some of the questions, black kids apparently want black kids as classmates, but not as friends* for example).

Also, I now have a better idea of what exactly 'self selected skin tone' is, its the skintone the kid wants to have.  There's a bias at play here, the things you want always seem better than the things you don't want (it occurs to me I know next to nothing about the ontology of biases, this may not be present in children).  It seems a strange way to manipulate the data too.

What really bugs me now is that half the questions are apparently about what adults or teachers think.  Questions about what a person believes and what a person thinks other people believe get different answers.  People are both more able to recognize bias in others, and have a bias against admitting to traits seen as negative.  Its possible this doesn't apply to children, but this is still sloppy.

Also, to address an earlier point of Ratatosk's, a number of the kids actually did refuse to answer the questions, sometimes as many as half the kids for a given question.

*Hyperbole.

YAY! More Data = More sense... I hate the 24 hour news cycle and their reporting of bits and pieces of useless drivel. Thanks for the update Requia.

[Golden Applesauce]

you've gotta be really careful when asking young children questions, because they don't answer rationally. This is, oddly, something that experimental psychology has only recently figured out.

For example, you know when they give the kid a doll and tell the kid "point to where the bad man touched you"?

Well it turns out that if you give a kid a doll with orifices and genitals which are not present on any doll the kid has ever seen, and you ask them just about anything, "point to the doll's nose", the kid is going to point to the doll's mystery hole or wee carrot or barn hole.

This - it's amazing that sociology/psychology as a field doesn't do the same level of method and model validation as other sciences.  I've been going through a lot of old chemistry/medicine notebooks lately, and those scientists spend at least a thousand pages of documentation each on answering questions like "Is the amount of Drug X (or its metabolites) in the rat's urine a good indicator of the amount of Drug X in the rat's blood plasma?" and even "Which analytical techniques give accurate measures of the amounts of Drug X in this animal bodily fluid?" repeated for every combination of drug/metabolite, test animal, and kind of bodily fluid.  In sociology/psychology it seems like all you need to do is come up with a questionnaire and run with it.  (The important tests - the ones with somebody's name and the word "Scale" in the title - do get validation.  I'm thinking more about the ones that get used in one study and then never again, while the conclusions drawn from it [which are not the same as the results] get endlessly parroted across news media, popular psychology books, and textbooks.)

For this study, I would have liked to see nonsense/control type questions among those asked.
Show me the child who has the most dinosaur toys.
Show me the tallest child.
Show me the child who eats the most carrots.
Show me the child who is really a space alien.
Show me the child with a red bicycle.
Show me the child with a blue bicycle.
Show me the child who has a pet turtle.

[Requia ☣]

Keep in mind that the people running this weren't sociologists or psychologists, they were CNN employees.

Though, they were bright enough to not put their names on the paper, so maybe they aren't a complete loss.