|
Always best when our beliefs are confirmed. It gives us a nice, cozy feeling when critical thinking skills curl up, pur, and take a nap on our collective lap.
Of course, the assumptions only hold if people wildly take correlation to mean causation and limit their variables to what they already believe must be the case. It just takes a few anecdotes for our bias to be confirmed, then zzzzz....
For example, take the claim that economic achievement is the overwhelming factor. For the article to work, it is an independent variable and must precede academic achievement. There's a prediction there, whatever the anecdotes may say.
The prediction is that in the '50s-'70s, when there were good union jobs paying high wages, the children of those workers should have had the same academic achievement as kids from white-collar families. The prediction fails: While the children of those parents achieved a lot in high school and went on to college at rates higher than the children of the poor, college admission rates for white-collar families was far, far higher and the drop-out rate for blue-collar kids who did go to college was far higher. If you disaggregate the data there are dominant and minor trends. One confound was the GI Bill, for instance, or the desire to avoid the draft; it still didn't allow a majority of the country, by SES, to send enough kids to college to constitute a majority of the undergraduate population. A case of anecdotes swamping the data, at least in our perceptions.
Making life harder for data geeks is that a lot of education and sociology research is advocacy based.
While family income is important, parental education levels swamp the contribution of income alone. When you look at the children of highly educated yet poor people--the stereotypical English major become secretary--you find that the parents' education still counts more than income. Moreover, typically family income is dependent on education levels. In other words, education as an independent factor accounts for a lot more of the variance than income. It nicely agrees with the result you get when you merely reason it through; it's good for the numbers to support the logic.
Even with all the hand-holding programs for first-generation college kids the correlation's still fairly strong--and the regression shows that family educational achievement is still the dominant factor. It's true in first grade, fifth grade, 8th grade, and in high school. Educated parents' get their kids to advance from .2 to .3 of a grade level during the summer break, according to some studies that looked at elementary school kids; the same studies show that the kids of parents in the bottom 25% SES *lose* that amount of "grade level." That's not "disruption to the learning process" caused by poverty; that's undoing the learning process because there's no support for what's learned in the household.
Of course income still has a contribution. When the kid's hungry, always moving, really stressed over basic necessities (or so overwhelmed by envy) that he can't think straight, education suffers. Even for most of the poor kids I've seen it's just not that bad. Achievement suffers more when they give up, though, because they've been so convinced that their parents' income and race determine their educational future. You want to see achievement increase long-term? Get the kids to believe that their effort will make a difference. (Then tell the rest of us how to do it.)
Decent income also allows the kids to think that if they do well in high school they might be able to afford college. Or it can move the kid out of a school or neighborhood predisposed to low educational achievement. Or even let the kid interact with high-achieving "peers." But in this it's still not the real variable that's being manipulated.
|