Schools do a poor job of diagnosing reading disabilities in children -- seeing the problem in boys who do not have it and missing it in girls who do.

According to another study published in JAMA, schools are up to four times more likely to label boys than girls as reading-disabled. But when standard research criteria are used to diagnose reading disability, the problem is found to occur about equally in both sexes.

The study, by researchers at Yale University and the University of Texas at Austin, evaluated 445 Connecticut schoolchildren whose intelligence, behavior and academic progress were regularly assessed from kindergarten through the third grade. Measures included a standard IQ test, reading and mathematics tests, and an inventory by teachers of classroom performance.

When reading disability was defined as a discrepancy between IQ and performance on the reading test, it was found in about 9 percent of second- and third-grade boys and in 6 to 7 percent of girls -- not a statistically significant difference. But teachers identified 10 to 14 percent of boys in those grades as reading-disabled, compared with only 3 to 4 percent of girls.

Less than half of the children labeled by schools as reading-disabled -- 13 out of 29 -- met objective criteria for the diagnosis.

Children whom schools considered reading-disabled but whose tests did not confirm the diagnosis were significantly more likely than average to show behavior problems in the classroom. Conversely, children whose tests indicated a reading disability but who had been "missed" by schools were significantly less likely than average to be classroom troublemakers.

The researchers reported that teachers tended to report more problems for boys than for girls in virtually every area.