By Alex Distefano
By Scott Snowden
By Anna Merlan
By Steve Almond
By Jena Ardell
By Jon Campbell
By Alan Scherstuhl
By Tessa Stuart
Ortiz's daughter, like Luke Serrano, scored well on the Bracken, but her OLSAT score, which counts for twice as much in the Department of Education's gifted calculus, pulled her down below the qualifying mark. In fact, visit any parent discussion board and you'll find parents with tales of otherwise academically savvy children who scored high-90s on the Bracken, which mostly measures such things as shape and color recognition, yet barely cracked double digits on the OLSAT, which specializes in "one of these things are not like the other" logic puzzles.
Harcourt Assessment, the Texas publishing giant that supplies both tests, insists they're proven to have sufficient validity (predicting future academic performance) and reliability (students' scores are consistent from test to test). An evaluation of the OLSAT by the University of Nebraska's Buros Institute, though, noted that its designers had provided no evidence that students who do well on the test go on to do well in gifted programs. Rather, they only showed a correlation between doing well on the OLSAT and on other tests—including previous versions of the OLSAT itself.
This is a recognized problem with standardized tests as a whole, says Robert Schaeffer of the National Center for Fair & Open Testing, which advocates against what it considers the misuse of standardized tests. "Indeed, test scores are good predictors of test scores," observes Schaeffer. "But how well it correlates with more significant outcomes like grades, graduation rates, going on to college, and longer-term adult performance is unknown, on almost all of them." The Buros study ultimately concluded that the OLSAT was appropriate for a "limited role" in screening students for special classes, as "one of a variety of instruments that could be used."
Adria Quinones, Manhattan mom of a G&T fourth-grader (and daughter of former schools chancellor Nathan Quinones), says her son's brightest classmate was admitted only by teacher recommendation, not test scores. "The kid's scary smart—you don't have to be around the child more than a few minutes before you say, 'This is a child who should be in a G&T program.' " Testing four-year-olds, to her mind, is ridiculous: "They poop at the wrong hour and boom, that's 20 points off their IQ."
The Department of Education says universal testing is fairer to poor and immigrant parents who may not have the savvy to sweet-talk human evaluators. But the numbers so far don't bear that out. Last month, the Times reported that the number of gifted seats going to students in the city's four wealthiest districts (the Upper East Side, the Upper West Side, Staten Island, and northeast Queens) had soared under the new system, from 24.9 percent to 39.2 percent.
Jacob counters that the number of kids scoring a 90 or better rose more in low-income districts (121 percent) than in other districts (60 percent). (Tests were given to entering students last year, but not used as an admission standard.) But the numbers show that's mostly because new outreach programs got more kids to test—as much as eight times as many in some districts that had few G&T programs in the past. A Voice analysis of Department of Education documents shows there are still dramatic disparities in results: In Manhattan's District 3, covering the Upper West Side, 35.8 percent of kids tested qualified for G&T programs. In District 9 in the South Bronx, meanwhile, the number of kids tested soared from 49 to 390, but only 2.8 percent—just 11 kids out of 390 applicants—made the cut.
It's figures like these that are only likely to harden the resolve of those who think that standardized tests are inherently biased against underprivileged kids. Children from more well-off families, testing critics note, are more likely to have had the kind of experiences—from enriched preschool classes to frequent library visits— that can help them at test time.
"It's a good thing that they're screening all kids for gifted and talented," says Schaeffer. "But the tool they're using is a skewed measure that only selects for a particular type of test-taking talent. And it ends up in the bleaching of gifted and talented classrooms."
The city currently says it will re- evaluate its standards for next year. The tests, though, are likely to remain, especially given that the Department of Education is in the midst of a five-year, $5 million contract with Harcourt to provide testing materials. (The department says it spent an additional $2.4 million this year on conducting and grading the tests.) Silverman suggests that at the very least, taking the top-scoring kids on either test, rather than using a weighted average, would help students who excel, say, in artistic areas without testing well on more left-brain skills.
To Barratt, meanwhile, the G&T wars are ultimately a distraction. "The real story is that there are not enough good schools for kids," she says. "So we have to lie, cheat, and steal spaces from each other to have any shot at a good education."