Ever since No Child Left Behind became law, a common complaint among
elementary school principals is that transfer students drag down test
scores, and they’re unfairly held accountable for the performance of
children who arrive mid-year.
I heard something like this just two weeks ago from a principal in
Humboldt Park, who told me that 15 new students had arrived that
week—just days before the ISAT.
Any truth to these gripes? The answer is both yes and no.
Ever since No Child Left Behind became law, a common complaint among elementary school principals is that transfer students drag down test scores, and they’re unfairly held accountable for the performance of children who arrive mid-year.
I heard something like this just two weeks ago from a principal in Humboldt Park, who told me that 15 new students had arrived that week—just days before the ISAT.
Any truth to these gripes? The answer is both yes and no. Transfer students do pull scores down, but not much, less than many would have you believe. Principals are held accountable for these scores by CPS, but not by the feds or the state.
Here’s what I found:
Districtwide, the average student mobility rate is 22 percent; at the school I mentioned in Humboldt Park, the rate is much higher. More than one in three students transfer in or out during the course of the school year.
No doubt mobility can have a tremendous negative impact on students, teachers and learning, especially when a third, and sometimes more than half, of the student population is unstable. But a close look at test score data shows that mobility doesn’t seem to have as big an impact on ISAT test scores as many imagine.
I compared school-level scores for all students, including those who transferred in mid-year and during the summer, to scores just for those students who had been enrolled in their schools by May 1 of the previous school year—the most stable population of kids.
On average, there was no difference, although schools with high mobility (more than a third) saw a difference of about 1 percentage point. (For this analysis, I took out schools with more than 30 percent English-language learners. I will explain why later.)
At the Humboldt Park school, for example, reading scores are 2 percentage points higher and math is about 4 points better when transfer students are taken out.
Which scores really count
State education officials do not count transfer students in determining whether schools meet average yearly progress—the benchmark set by No Child Left Behind policy.
Before 2006, state education officials counted scores for students who were enrolled in their schools by Sept. 30. But teachers and principals complained that it wasn’t fair to hold them accountable for students who were enrolled in their schools only five months, says Illinois State Board of Education spokeswoman Mary Fergus.
According to the Consortium on Chicago School Research, last year about 23 percent of CPS students were not enrolled in their home schools the previous May 1.
By contrast, Chicago’s school probation policy calls for all students sitting in a class on test day to be counted toward a school’s averages.
“These are our kids and the schools have to be accountable to them,” says Ginger Reynolds, head of CPS’ Department of Research, Evaluation and Accountability. Discounting test scores for transfer students was not considered, she says. “We just didn’t think it was the thing to do.”
Yet, the district did think it was right to exclude test scores for children who are learning to speak English. While state officials count scores for all English-language learners, who for the first time this year were required to take the regular form of the ISAT, CPS objected, and counted only those students who had been in English language programs for six years or more.
When ELL students in those schools with more than one-third ELL populations are excluded, CPS schools perform, on average, 8 percentage points better in reading and 4 points better in math.
There are fewer CPS schools on probation (206) than on the list of those that didn’t make average yearly progress under NCLB. (383). That’s because the district’s complicated accountability policy provides for a targeted range of test scores schools must hit.
Currently, the bar to meet AYP is fairly high, with schools needing 62.5 percent of students meeting standards. It will go up to 70 percent next year.
—Sarah Karp