Questions raised on MCAS data
Many students did not list their racial identity
The narrowing of the racial achievement gap on the MCAS test, hailed by Governor Mitt Romney and other top officials last week, has been thrown into question because almost 38,000 students omitted their race on the exam, the state Department of Education has disclosed.
In what state Education Commissioner David P. Driscoll termed a mystery, about 10,700 10th-graders -- and about 27,000 in other grades -- did not indicate their race when they took the 2003 test. While their performance was included in the overall number of students who passed, they were not counted in the breakdown of scores by race.
The students were not required to supply the information, but the omissions raise concerns about the DOE's published results by race -- the scores that Romney and Driscoll touted as evidence of the narrowing of the achievement gap.
Driscoll yesterday cautioned that he does not know whether the inclusion of more students' scores will change the results by race. Typically, about 1 to 3 percent of test-takers do not indicate their race, but this year, the number was more than 7 percent. About 527,000 students in grades three, four, five, six, seven, eight, and 10 took one or more sections of the MCAS in April and May, in English, math, or science.
"Maybe the kids finally said, `Hey, look, it's none of their business who knows this,' " Driscoll said. "We're going to have to look at it. It's a very interesting phenomenon."
Including the results in the racial breakdowns will not change the overall grade-by-grade scores -- including the fact that 75 percent of 10th-graders passed the MCAS test on their first try. Passing the English and math sections of the Massachusetts Comprehensive Assessment System test became a graduation requirement beginning with the class of 2003.
But the added students could rearrange achievement when broken down by race, a measure that state leaders use to determine whether black and Latino students are catching up to their white and Asian peers.
"Will the percentages change? They may change a little bit," said Jeff Nellhaus, associate commissioner for student assessment. For example, 84 percent of white 10th-graders in the class of 2005 passed MCAS on their first try, compared with 44 percent of Latinos and 52 percent of blacks. In 2001, 77 percent of white 10th-graders passed on their first try, compared with 29 percent of Latinos and 37 percent of blacks.
The drop in the number of test-takers identified by race this year compared with last year was brought to DOE's attention by a Globe reporter. Officials said they did not check the differences because they wanted to release the MCAS scores as soon as possible.
Department officials now will match the names of students who did not pick a race with their racial information contained in a new database that has basic information on every public school student. Revised scores by race will be released in about two weeks.
But some specialists said the issue stretches beyond Massachusetts. The federal No Child Left Behind Act requires schools to report achievement by race, and imposes penalties if performance does not improve. Schools have to combine those goals with students' increasing rejection of race labels, said Stephen Sireci, an education professor at the University of Massachusetts at Amherst.
"You want to respect the privacy of individual students and students' families. At the same time, you want actual mechanisms in place to make sure minority groups are making the gains we're striving for," said Sireci, codirector of UMass-Amherst's Center for Educational Assessment. Students can choose to identify themselves from six categories: Asian, black, Hispanic, mixed/other, Native American, and white.
Students' refusal to indicate race did not surprise Randolph Superintendent of Schools Arthur J. Melia, who has talked to youths in his district who ignored the race question on the test. Randolph schools are among the most diverse in Massachusetts, with 41 percent black students, 38 percent white, 14 percent Asian, and 7 percent Latino.
"Some of them I suspect are of mixed race. Some protest the categorizing of kids into different groups like that, not wanting to contribute negatively to one group, or positively, whatever the case may be," Melia said.
Nellhaus said the validity of the MCAS test remains strong. But MCAS critics, such as Boston College education professor Walt Haney, said the discrepancy boosts the contention that the test is unreliable.
"They're getting caught in a situation where they're putting out results that, once you dig, appear to be highly questionable," said Haney, senior research associate in the school's Center for the Study of Testing, Evaluation and Educational Policy.
© Copyright 2003 Globe Newspaper Company.