"Maxwell's demon is a thought experiment created by the physicist James Clerk Maxwell in 1867 in which he suggested how the second law of thermodynamics might hypothetically be violated.In the thought experiment, a demon controls a small door between two compartments of gas. As individual gas molecules approach the door, the demon quickly opens and shuts the door so that only fast molecules are passed into one of the chambers, while only slow molecules are passed into the other. Because faster molecules are hotter, the demon's behaviour causes one chamber to warm up and the other to cool down, thereby decreasing entropy and violating the second law of thermodynamics."
I.
In my review of The Third Pillar, I quoted a reference in the book to a study about schools' gifted student programs. The study found that "there is no overall improvement in results as the benefit to the brightest is cancelled out by the drag on the rest." In other words, putting a student in a gifted program will raise their scores while also lowering the scores of the students in the class the gifted student was removed from.
In a study called The Social Structure of Schooling, researchers found a similar phenomenon when schools used tracking. Put a bunch a smart kids together and their grades go up. Put a bunch of low performing kids together and their grades go down.
This is Maxwell's Demon in the classroom.
There should be more disorder, more entropy, more poor kids finding success and rich kids falling behind. But there is a demon in our society that interferes, which causes feedback loops we all get stuck in.
You can draw different conclusions from these studies. Maybe you think it means that successful people are only that way because of unearned privilege of being placed in the fast molecule room, or that schools need to spend more on low performing students at the expense of high performing students. But I think the lesson should be that society underrates the impact of peer influence.
The number one villain here is zoning regulations, not a lack of spending on public schools in poor neighborhoods. Mark Zuckerberg's $100 million donation to Newark, NJ public schools didn't work because it did not address the core problem: kids being stuck in the slow molecule room. Infusing the slow molecule room with money isn't the same as infusing it with fast-molecule students.
Instead, we need to kill the demon and [Ronald Reagan voice] tear down that wall.
II.
It isn't just poor parents without college degrees that are holding their kids back. In The Nurture Assumption, Judith Harris writes:
"British studies have shown that when delinquent London boys move out of their city their delinquency rates decline - even if they move with their families. By living in one neighborhood rather than another, parents can raise or lower the chances that their children will commit crimes, drop out of school, use drugs, or get pregnant."
It's probably wrong to blame the family for a problematic child, often times it's the school and community that influence him.
A New York Times study shows that poor white boys are more likely to become successful than poor black boys, and that rich white boys are more likely to stay rich, while rich black boys are more likely to end up poor, which is disheartening to say the least.
The NYT study looked for outliers and found that:
"The few neighborhoods (where poor black boys do well as well as whites) ... were the places where many lower-income black children had fathers at home. Poor black boys did well in such places, whether their own fathers were present or not.
"Other fathers in the community can provide boys with role models and mentors, researchers say, and their presence may indicate other neighborhood factors that benefit families, like lower incarceration rates and better job opportunities."
This is perfectly in line with Harris's research: It's not the makeup of the child's family that matters, it's the makeup of the families in the child's community.
I've touted the importance and privilege of two-parent families, but I wasn't totally correct. Outside of the home, parents have little influence over their children. Instead, children absorb the values of their peers, which is partially a reflection of their peers' parents.
The best predictor of a child's future earnings is a combination of their parents' earnings and their friends' parents' earnings. So when towns pass things like single family zoning, they prevent the benefit a disadvantaged child would get from the positive influence of living in that community.
This is Maxwell's Demon in the community.
III.
Here is the problem with tearing down the wall and creating a demon-free society. If you eliminate gifted programs and other opportunities for high income parents to give their kids a leg up, they will just find new ones. The Seattle public schools equity initiative led to high income parents pulling their kids out and putting them in private schools. And even if highly educated parents to opt to mix their kids with low-income students, their children will see a reduction in test scores and an increase in things like dropping out, incarceration, or teen pregnancy (offset by a decrease in those things for low-income children).
The challenge is to create a great enough incentive for high income parents to open the door and let their kids into the slow molecule room and to tweak the settings just right to maximize the benefit of the slow molecules and minimize the harm of the fast molecules. There are a few uphill battles. First, scalability.
Research shows that children learn better in smaller classrooms.
"After four years, it was clear that smaller classes did bring substantial improvement in early learning in cognitive subjects such as reading and arithmetic. Following the groups further, the Lasting Benefits Study demonstrated that the positive effects persisted into grades 4, 5, 6, and 7, so that students who had originally been enrolled in smaller classes continued to perform better than their grademates who had started in larger classes."
In small classes, teachers are able to convince students they are a unit so the salience becomes the classroom, rather than race or class distinctions, which have the Maxwell's Demon effect of feedback loops. I mentioned earlier that spending more on education will not fix the problem, but the right kind of spending can help. Any spending to reduce classroom size is a good start.
Second, not just any teacher will do. You need the right classroom leader who can make the classroom unit salient. A leader will define the stereotype the classroom has of itself and prevent students from falling into separate groups that perpetuate the Maxwell's Demon effect. The right teacher, like Miss A, can turn the entire class into an "us" that sees itself as scholars.
Third, you have to get the mix right when you try to de-sort the molecules. Without the right things in place, mixing molecules can make things worse. Janet Schofield's research in the racially-diverse Wexler school showed that students clustered by race and adopted opposing attitudes toward school that became more pronounced over time.
Harris's research found that:
"Whether a classroom of kids will split up into contrasting groups depends partly on how man kids there are ... whether kids will form groups that differ in village of origin, or in race, ethnicity, religion, socioeconomic class, or academic ability, depends on how many there are in these social categories....
Number is important. A few students from a different socioeconomic class, ethnic group, or national background will be assimilated to the majority, but if there are enough of them to form their own group they are likely to remain different and contrast effects may cause the differences to increase. At intermediate numbers, things can go either way: two classes with the same number of majority and minority students may in one case split up into groups and in the other remain united. It will depend on chance events, on the characteristics of the individual children, and, crucially, on the teacher."
The final challenge is stereotype threat. Claude Steele found that all you have to do to lower the score of a bright black kid on a test of academic ability is to give her, before she takes the test, a short questionnaire that includes the question 'Race?' The same thing happens when you test a woman and ask her gender. (There is some evidence of replication failure for stereotype threat.)
In contexts where gender is less salient, girls and young women do better in science and math. Women's colleges produce a disproportionate number of outstanding female scientists. Likewise historically black colleges and universities produce a disproportionate number of outstanding black scientists.
This research is a good argument for HBCUs, and unfortunately, a bad argument for black spaces in majority-white colleges.
Unfortunately we don't know the right number yet to know the ideal settings for getting the most benefit out of peer influence and the best ways to minimize the Maxwell's Demon effect. But if I could wave a magic wand I would do three things.
Reduce classroom size. Reward/retain/attract teachers who are leaders in the classroom.
Abolish single family zoning.
Incentivize highly educated, high income parents to put their kids in low performing schools.
No comments:
Post a Comment