You may have heard about the recent study by Mark Kishiyama et.al. described by USA Today as follows:
A new study finds that certain brain functions of some low-income 9-
and 10-year-olds pale in comparison with those of wealthy children and
that the difference is almost equivalent to the damage from a stroke.
Here is more detail:
…[they] rigged up the noggins of 26 kids — with an average age of 9.5 years –
with probes that sense the ebb and flow of electrical current in
different regions of the brain. Then, they put them through a battery
of neuropsychological tests. Half of the kids came from families with
annual incomes that averaged just over $27,000 and generally had low
levels of parental education; the other half came from families where a
primary caregiver had completed at least four years of college and in
which annual household income averaged a little more than $97,000.
The paper is here. Who better to ask about this than Michelle Dawson? Michelle wrote to me:
I read the poor vs rich
kids brains study (Kishiyama et al.). It’s a very small study (13 in
each group) and the groups aren’t matched on ethnicity. In the major
task (the one which got media attention), where the authors looked at
ERPs [TC: here is a link on ERP], the performance of the two groups was the same. The performance
of the two groups on a Stroop task, a classic test of what the poor
kids are said to be incapable of, was also the same. The major
performance difference between groups was on vocabulary (the WISC-III
vocabulary test), but only a few tests were used. There was no attempt
to match the groups on IQ.
Just to repeat two key points: a) the observed difference in electrical current patterns may depend on IQ differences, not poverty, and b) on the actual major task the poor kids did just as well. There are tasks where the poor children do less well but this is hardly news.
Popular science reporting on neuro issues is very often not to be trusted.
Addendum: There is more from Michelle Dawson in the comments.
I’m just boggling at seeing someone with my income described as “rich” and “wealthy”. I mean, maybe by global standards, but by US standards we have a modest house in a dodgy neighborhood, etc.
If two 5 person groups is all you need in a study, you must be expecting very large differences between the groups to get anything of meaning. A priori I wouldn’t expect a very large difference between these two groups even though they had 13 people per group.
I’m with James. Particularly with a study that is both controversial and has broad policy implications, you would want to be conservative in your conclusions.
Note this shows up in USA Today BEFORE it is published, which probably means somebody’s PR apparatus is involved.
Vocabulary testing is particularly likely to be subject to environmental factors. [e.g. I'd do very poorly on a Spanish, Polish, Chinese, Hindi, Korean or even Appalachian vocabulary test]
This particular study may have problems, but it would be surprising if the basic result turned out not to be correct — e.g. that there are frontal lobe mediated attentional/executive functioning differences between rich and poor children.
We already know that socioeconomic status is highly correlated with academic success. And we also know that there is a correlation between differences in executive functioning (impulse control, ability to delay gratification) at very early ages and later academic success, with Michel’s ‘Marshmallow Tests’:
http://econlog.econlib.org/archives/2007/10/stanford_marshm.html
But if such differences exists, it’s an unjustified leap to assume that money, per se, is the critical issue, and even more that any such differences could be erased by intensive pre-school interventions. There may be a significant genetic component, and there may also be family culture factors that intensive interventions cannot make up for.
The issue with sample size is both one of statistical power and generalizability. If they wish to argue the null, then statistical power becomes relevant. If they have statistically significant results, the question becomes one of generalizability. While this is an issue with studies of any size less than the entire population, small samples comparisons are very difficult to control for many other relevant dimensions.
Now, I have alot of respect for the authors, and the journal (J of Cog Neuroscience) has a fairly tough review process so I’m happy to *believe* the results. Performance/ERP relations are complicated so a priori I don’t necessarily discount ERP differences if they are not associated with performance differences although it suggests some complexity to the interpretation.
Poverty is associated with pervasively increased levels of stress, reduced environmental stimulation, poorer nutrition, and on a day to day basis, less sleep (yep, we can talk about how overworked everyone is but the poor actually sleep less well than the comparatively well off. Sorry but I don’t have the reference handy). But, its also possible that the poor participants experienced more stress, for example, at the time of testing. Note that some of these effects are very transient (e.g., sleep might be), while others may be much more long lasting (nutrition, environmental stimulation, situational stress). The why is the completely unclear then.
Hmmm… so, kids of grad students (average salary 20k) are dumber than kids of union auto workers. Popsci reporting sucks.
If they reported ethnic differences they would be called racists.
Yeah, no social scientists ever study race, it’s such a taboo topic! That’s why no one no one knows if there’s a black-white test-score gap, or if there’s a black-white wage gap! POLITICAL CORRECTNESS!
Re: IQ vs. poverty; when Tyler says, “may depend on IQ differences, not poverty” this seems to indicate that he believes IQ is the type of thing that can *cause* differences in brain patterns.
One way that majoring in economics (and consequently taking econometrics) has changed my life is my automatic disbelief of all popularly reported scientific findings.
I’m pretty sure my friends are tired of my normal response to these things: “Meh… I want to see their methodology.” That, and constant complaints of endogeneity.
I believe it all starts with prenatal care and nutrition. Good organic food or even fresh fruit is expensive. Transportation to prenatal care, insurance issues, etc. leave the poor a little bit more vunerable. I say brain development would definately be affected by nutrition, prenatal care, and early postnatal care.
Even the substantive claims of the study (about the equivalence of poverty and specific brian damage) are pretty far-fetched considering the shortage of evidence, and it is really only speculating about “stress” and lack of cognitive stimulation. There’s no recipe in there for finding a cause even if the effect is real, which there is reason to doubt. I’m perfectly prepared to believe that poverty and retarded intellectual development are associated, but I’m being given no reason to accept such a neurological phenomenon in general exists in general, let alone that it is part of the mechanism of such association.
So where do the nutrition theorists and critics of intellectually pompous capitalist economy come from? (Europeans reading this will be very surprised by the latter, since it is a common assumption of the intelligentsia on this side of the Atlantic that US capitalist culture denigrates the intellectual.) We have no idea what the kids ate.
A poor person who makes $27,000 year. I served in the military for 23 years and I was happy when I could make over $10,000! We had two children to raise and of course feed and whatever additional expenses. Also, the military did not want people thinking that its members were not paid well so they made sure that none of us were able to even qualify for food stamps or other low income assistance. I guess all those years and separations, wars and hard times were just a bad dream. I’m sure glad I woke up.
Even if the groups were large enough to constitute a significant study, the groups were not matched equally. Compare the brains of children living in poverty who have parents with college degrees to wealthy children of parents with similar degrees.
To all of you from a comfortable background: You cannot begin to imagine the differences between living in a household with less than a $30,000/year income if you earn $100,000/year. You may as well try to imagine having an I.Q. of 40. Children raised in poverty-stricken families have much bigger worries than can be defined by standardized test. Are these researchers aware of the correlation between thought patterns and brainwaves? Thought patterns of poor children are different because they have different thoughts. I shouldn’t need to spell this out for anyone. I’m poor; what do I know?
Specifically to the person who mentioned prenatal care: Prenatal care is a cause of stress to low-income women, particularly those without health insurance. Factor in the stress of the hours spent at your local Dept. of Human Resources, and the jumping through hoops necessary to receive Medicaid, and you have impacted the unborn child in ways which known to cause The most harm. If we are accepting the validity of a scientific study using only 26 children, we may as well accept my personal case-study evidence with the same validity. My youngest son, who was born without the so-called benefit of prenatal care, would assuredly outscore you on any I.Q. test.
This is your brain.
This is your brain on poverty.
Did the study control for lead exposure? Lead is a neurotoxin, and exposure to it is associated with poverty. If the cognitive differences were triggered by such an environmental cause, genetics and/or parenting or other sociological phenomena would not underlie the differences detected in the study. I might add that the Bush White House recently interfered with the EPA in order to exempt many urban lead pollution sources from monitoring.
BTW: John W., at Dec 12, 2008 12:07:16 PM, brought up social Darwinism, but he seems to have misinterpreted the statement that “the observed difference in electrical current patterns may depend on IQ differences, not poverty.† I would note that there is nothing in the original post that suggests that poverty is caused by low IQ or genetics, and so there is nothing that appears to advocate a social Darwinist position; John W.’s apparently willful misinterpretation seems ideological to me †¦
Comments on this entry are closed.