From the New York Times:
Why Talented Black and Hispanic Students Can Go Undiscovered
By SUSAN DYNARSKI APRIL 8, 2016
Public schools are increasingly filled with black and Hispanic students, but the children identified as “gifted” in those schools are overwhelmingly white and Asian.
The numbers are startling. Black third graders are half as likely as whites to be included in programs for the gifted, and the deficit is nearly as large for Hispanics, according to work by two Vanderbilt researchers, Jason Grissom and Christopher Redding.
New evidence indicates that schools have contributed to these disparities by underestimating the potential of black and Hispanic children. But that can change: When one large school district in Florida altered how it screened children, the number of black and Hispanic children identified as gifted doubled.
That district is Broward County, which includes Fort Lauderdale and has one of the largest and most diverse student populations in the country. More than half of its students are black or Hispanic, and a similar proportion are from low-income families. Yet, as of 10 years ago, just 28 percent of the third graders who were identified as gifted were black or Hispanic.
In 2005, in an effort to reduce that disparity, Broward County introduced a universal screening program, requiring that all second graders take a short nonverbal test, with high scorers referred for I.Q. testing. Under the previous system, the district had relied on teachers and parents to make those referrals.
Several generations ago, Sir Cyril Burt was knighted by a British Labour government for introducing IQ testing into British schools, which roughly doubled the proportion of working class children achieving various honors over the previous more subjective systems. Also, Burt pointed out in 1912 that women and men appear to be surprisingly equal in IQ.
The economists David Card of the University of California, Berkeley, and Laura Giuliano of the University of Miami studied the effects of this policy shift. The results were striking.
The share of Hispanic children identified as gifted tripled, to 6 percent from 2 percent. The share of black children rose to 3 percent from 1 percent. For whites, the gain was more muted, to 8 percent from 6 percent.
We’ll come back to that.
Why did the new screening system find so many more gifted children, especially among blacks and Hispanics? It did not rely on teachers and parents to winnow students. The researchers found that teachers and parents were less likely to refer high-ability blacks and Hispanics, as well as children learning English as a second language, for I.Q. testing. The universal test leveled the playing field.
Multiple factors could be at work here: Teachers may have lower expectations for these children, and their parents may be unfamiliar with the process and the programs. Whatever the reason, the evidence indicates that relying on teachers and parents increases racial and ethnic disparities.
Uh, well, reading Card’s actual paper makes clear there was also another factor at work, although this doesn’t get mentioned in the NYT coverage.
From p. 2 of the Card – Giuliano paper cited in the article:
CAN UNIVERSAL SCREENING INCREASE THE REPRESENTATION OF LOW INCOME AND MINORITY STUDENTS IN GIFTED EDUCATION?
Working Paper 21519
… In response to these disparities the District introduced a universal screening program in spring 2005. Under this program, all second graders completed a non‐verbal ability test, and those scoring above a threshold of 130 points (for non‐disadvantaged students) or 115 points (for ELL [English Language Learner] and FRL [Free / Reduced Lunch] participants) were eligible for referral to a District psychologist for IQ testing.
Oh, so that’s it! They had a quota, a one standard deviation lower bar, for poor or immigrant students. Assuming a scoring system where 100 = the median and the standard deviation = 15, then “non-disadvantaged students” had to score at almost the 98th percentile while ELL and FRL students had to score only at the 84th percentile. Alternatively, the non-disadvantaged had to be in the top two percent while the disadvantaged only had to be in the top 16 percent.
Surprise, surprise …