It is not surprising that the editors of the Charlotte Observer published an editorial today criticizing the recent Supreme Court decision to limit school assignment by race. But instead of launching the usual liberal diatribe against racist conservatives, they attempted to pull off the “research says” (similar to Simon says) approach.
The move brings with it an air of objectivity and learnedness not typically associated with a McClatchy newspaper. Of course, they can’t pull it off, but it does produce a gem of a sentence, “This study reaffirms that school diversity is more than feel-good politics of liberals. [They should know.] It’s good public policy and an educational asset.”
Now to the research. A new study by UNC researchers concluded, “Minority segregation in elementary schools was associated with lower student reading performance after accounting for child and family background, classroom instruction, and school-level poverty.”
Had they read the entire study, the limitations of the study would have disappointed (but not deterred) the Observer editors. Some of the more important limitations include:
1. Teacher and school variables were obtained using questionnaires. Self-reporting is notoriously biased, and in this case, likely exaggerated the teacher quality measure. In other words, these teachers were probably not as competent as they reported they were in the questionnaire.
2. The report pointed out that the generalizability of the results may have been compromised because of student mobility and changes in student assignment. While the researchers eliminated movers, it appeared to ignore teacher or administrator turnover, which is higher than average in low-income schools. Higher turnover is associated with lower student performance.
3. The researchers do not (and cannot) claim that racial segregation causes lower reading scores. As the researchers point out, “this research is not a causal analysis of the determinants of reading development. Consequently, the potential for policy recommendations is limited.” It merely points out the race and performance variables were associated in a modest way.
4. Speaking of the race or segregation variable, we have no idea if this is actual segregation. The researchers say that segregation occurs when the “number of children from ethnic and racial minorities attending the school is greater than or equal to 75%.” Unfortunately, the researchers do not report the racial composition of the communities or school districts from which they took their sample. Therefore, it is possible that the school had a population of students comparable to the community or school district.
