Experts say it's impossible to know for certain why schools in the Los Angeles Unified School District posted the second-highest improvement in the country on the Trial Urban District Assessment ("What's driving L.A. Unified's better test scores?" Los Angeles Times, Dec. 20). I think they're asking the wrong question. Instead, they should be focusing on the causes of the achievement gap that continues largely unabated.
In fact, the gap persists in other urban districts as well. For example, in the District of Columbia, fourth-grade students from impoverished homes had an average math score that was 49 points lower than students from advantaged backgrounds. The gap was 21 points wider than it was in 2003. Eighth-grade students posted an average score in math that was 42 points lower than their advantaged classmates. The gap was 18 points wider than it was in 2003.
I suspect similar patterns exist in other urban districts because poverty plays such a powerful role in performance. I don't doubt for a second that some urban schools manage to shrink the gap by significant margins. They deserve lavish praise, but they are outliers. In other words, the exception doesn't disprove the rule.
It's interesting to note that at least in Los Angeles smaller class sizes have not mattered. In the Los Angeles Unified School District, test scores actually improved when class sizes increased as a result of the state's budget crisis. But I question whether this is true elsewhere. According to the Labor Department, public schools employ about 250,000 fewer people than before the recession, even though student enrollment has grown by more than 800,000 students.
Teachers know from experience that large classes make it harder to meet the individuals needs and interests of students ("Subtract Teachers, Add Pupils: Math of Today's Jammed Schools," The New York Times, Dec. 22). Whether this is reflected in TUDA or on standardized tests is beside the point because learning is not merely seen in test scores. It's also seen in non-cognitive outcomes. How can teachers possibly attend to this domain when their classes are overflowing?
I welcome measurements of school performance. But I urge caution in making inferences based solely on scores. Education is more complex than that.