Public school watch: Concerning the role of Advanced Placement classes!

TUESDAY, MAY 24, 2016

Lexington versus Detroit:
On May 3, Motoko Rich wrote a puzzling report in the New York Times about the state of the public schools.

She focused on a new Stanford study; the study compares academic achievement by public school students with a measure of their socioeconomic status. At one point, Rich tickled the keys of one of our top pet peeves.

In what follows, Rich makes an important point about a type of relative disadvantage faced by some middle-class black and Hispanic kids. In our view, she's also working from script in the highlighted passage:
RICH (5/3/16): What emerges clearly in the data is the extent to which race and class are inextricably linked, and how that connection is exacerbated in school settings.

Not only are black and Hispanic children more likely to grow up in poor families, but middle-class black and Hispanic children are also much more likely than poor white children to live in neighborhoods and attend schools with high concentrations of poor students.

These schools can face a myriad of challenges. They tend to have more difficulty recruiting and keeping the most skilled teachers, and classes are more likely to be disrupted by violent incidents or the emotional fallout from violence in the neighborhood. These schools often offer fewer high-level classes such as Advanced Placement courses, and the parents have fewer resources to raise extra money that can provide enhanced arts programs and facilities.
According to that passage, middle-class white kids are likely to attend schools with other middle-class students. Middle-class black and Hispanic kids are likely to attend schools with a larger proportion of low-income kids.

For middle-class or low-income kids, this represents a possible form of disadvantage. "Schools with high concentrations of poor kids" may well confer serious disadvantages on a wide range of good decent kids.

That said, we're always annoyed by the scripted passage in which the writer notes that low-income schools "often offer fewer high-level classes such as Advanced Placement courses."

That's almost surely true, of course. But it tends to be offered as a snarky criticism of those schools. This type of criticism makes little sense.

Should anyone be surprised when low-income schools offer fewer Advanced Placement courses (or "gifted and talented" classes)? Consider the profile of two systems from the first interactive graphic within Rich's report:
Average achievement levels, two school systems, grades 3-8:

Lexington, Mass.: 3.8 grade levels above average
Detroit: 2.3 grade levels below average

Achievement gap: 6.1 grade levels
Lexington is a high-SES suburb of Boston. Detroit is a large, low-income city which is struggling just to hang on.

The achievement gap between the two student populations is both huge and disastrous. In grades 3-8, the average kid in Lexington is 6.1 grade levels ahead of the average kid in Detroit, according to the metrics of the new study by Stanford.

That's a gigantic, disastrous gap. And remember, that's just a comparison of the average students in those two school systems:

Roughly half the kids in Lexington (grades 3-8) are more than 3.8 grade levels above average! Meanwhile, roughly half the kids in Detroit are more than 2.3 grade levels below average.

For those kids, the achievement gap is even wider than 6.1 grade levels. That represents a tragic societal disaster.

That said, should anyone be surprised if Lexington offers more "gifted and talented" courses in grades 3-8 than Detroit? If those districts offered the same number of such courses, wouldn't someone be committing educational malpractice somewhere?

The relative absence of Advanced Placement classes may well disadvantage the higher-achieving students in a low-achieving school district. But it's obvious why such districts offer fewer such courses and classes.

That said, you'll often see showboating journalists snark about this state of affairs. It's an easy way to play the game. It makes us liberals feel good.

Markers of socioeconomic status: According to the first New York Times graphic, Lexington's students come from families whose median family income is $163,000. The corresponding figure in Detroit is $27,000.

Remember, income is only one measure of SES used in the Stanford study. Family structure and parental educational attainment were also used by Stanford in assessing students' SES.

And no—Lexington isn't at or even near the top in median family income around the country. You can see this by clicking around in the Times' first graphic.

Nor is Detroit at the bottom in median family income. Click around; check out our struggling world.


  1. This is just a nitpick because I agree with Somerby's points here. An average achievement of 3.8 grades above average does not mean that roughly half of all students will be above that 3.8-above-average level. This is only true when 3.8-above-average is the median achievement. Somerby's statement that half the students will be above some score is only true when the distribution is normal. If the distribution is skewed by very high or very low scores, there can be many more scores in the tail of the distribution opposite the skew, than there are in the other tail and more than half of students will be in that direction and far fewer in the other direction. It is only when the distribution is symmetrical that the median and the mean scores are close to each other and his statement is true.

    With very large numbers of students in Detroit, the distribution is more likely to be normal. In a smaller city like Lexington, most students might be high achieving but there may be some exceptionally low scores (caused by disability rather than poverty) that skew the distribution and depress the mean, making it unrepresentative of most students. In schools, it is a few very high achieving students who skew the distribution and raise the mean above what is true for most students in that school.

    It would be better if Somerby said median when he means 50% are above/below, instead of mean or the less descriptive term "average."

  2. I agree with most of your comment, Anon, but would add one point. The Times report addresses "average", not "mean". I don't think one can tell from the report whether the word "average" is used here to mean "mean" or "median"

    Also, the Times has a usage that I dislike, namely, addressing a difference in averages as if it were a difference in all students. E.g., they write, "There are large differences between white students and their black and Hispanic classmates." This statement isn't quite correct. There are differences in the averages of these groups, but that difference in averages doesn't tell you anything about a particular white student or a particular black student.

  3. "Remember, income is only one measure of SES used in the Stanford study. Family structure and parental educational attainment were also used by Stanford in assessing students' SES."

    Yes indeed kiddos. And don't forget that those two measures, education level of adults and number of adults in today's modern houselhold are the two things that will do the most to drive up income levels in the first place. So pat Bob on the head for this self evident nonsense and tell him how cute he is when he calls you "dumb."

    1. Serious problems exist with our high-profile, influential national reporting on schools. Here, we see one typical example of that, yet shits like Anon12:01 always pretend to find a bigger problem with the few pointing that out.

    2. The one pointing out "serious problems" makes up his data and invents non existent rules. You call his critics "shits" for pointing that out.

      How are the trombone lessons coming along?

    3. "Serious problems" in scare quotes? Hahahahahahaha. Talk about showing your hand.