Study: Drops in Class Rank Affect Student Outcomes
District open-enrollment policies have gained traction in many urban areas as a way to give students more access to high-quality schools, but a new study suggests students who were "big fish" in their neighborhood schools can have a rough road in a new school, even if it is academically better.
Researchers at the University of Chicago Consortium on Chicago School Research found that class rank affects Windy City students' grades and test scores and even their persistence in college, and that students of similar academic ability in 8th grade can nonetheless hold very different class ranks depending on the school they choose to attend.
"Even if you are high-achieving, you are surrounded by other high-achieving kids, and somebody has to be the lower rank," said Lauren Sartain, a research analyst at the Chicago consortium, during a presentation of the forthcoming study at the annual meeting of the Society for Research on Educational Effectiveness here.
The study grew out of a puzzling finding, both in ongoing Chicago research and in other open-enrollment districts such as New York: While open enrollment often allowed students to move to more academically rigorous schools, there were surprisingly few academic benefits for students attending the most elite, selective public high schools. That spurred the researchers to dig into what happened when students attended high schools that were different from those they had been assigned by neighborhood.
Sartain and her colleagues tracked four groups of Chicago students entering 9th grade from 2008 to 2011, totaling more than 72,000 students. More than three out of four students in the district attends a different high school from his or her neighborhood-assigned school.
They calculated each student's general academic ability based on his or her grade point average, attendance rate, and the number of suspensions accumulated in 8th grade, as well as math and reading test scores for grades 3-8, and, to a lesser extent, the student's own reported study skills and habits. Separately, the researchers calculated each student's class rank as the schools themselves do, based on their 8th grade standardized test scores. They compared the class ranks for students at both their neighborhood-assigned school and the one they actually attended.
While average-performing students remained at about the same rank or even moved up a little in their new schools, low-performing students dropped to even lower ranks, and the students in the top 20 percent of their class in 8th grade showed an even bigger drop in class rank when they moved to high schools of choice.
No significant differences were found in the effect of class rank for male versus female students or students of different races.
Students who started at the same ability level at the end of 8th grade ended up with a full standard deviation difference or more in their class rank at their next school, depending on which school they attended.
"Think about a student who has been at the top of their class for eight years, basically," said Sartain, noting that most Chicago primary schools follow a K-8 model rather than having separate middle schools. "You're really used to being the big dog. We think there are definitely going to be psychological effects, for, 'I used to be at the high end, but now I'm in high school and I'm in the middle of the pack.' "
There were more concrete problems than just a hit to students' egos. Students of the same ability level who entered high school with a class rank in the third quintile—right at average—had GPAs that were .04 points higher than students in the bottom 20 percent, and .06 points lower than students ranked in the top 20 percent at their school. There was a similar, statistically significant gap for state test scores.
Impact on College Success
RAND researcher Rebecca Herman, who was part of the SREE symposium but was not associated with the study, said the differences in outcomes could stem from tracking students—for example, recommending that freshman take honors classes only if they were in the top 20 percent of their incoming class—but she doubted individual teachers treated students differently based on their class rank.
"Most teachers are not in their head making the same ranking [researchers] made," she said at the SREE symposium. "The filters that teachers and staff have when they are looking at the data is critical."
Interestingly, class rank did not affect whether students chose to enroll in college after graduation, but it did affect whether they stayed in college. Based on the two years of data on students who entered college during the study, Sartain and her colleagues found that students who had been in the top 20 percent of both rank and ability level in 8th grade but were ranked only average in their high school class were .5 percentage points less likely to finish their freshman year in college than students who continued to be ranked in the top 20 percent.
That is troubling, Sartain said, since one of the reasons parents often give for enrolling students in higher-performing high schools is to improve their chance of being accepted to good colleges.
The researchers are now digging into the data more deeply, to figure out what role tracking, personal identity, peer connections, and other issues might play in students' trajectories at a new school.
The results suggest school leaders and parents should take a more holistic look at what schools offer their students, particularly during the high school transition period, said study co-author Marisa de la Torre, the associate director for professional development at the Chicago consortium.
"We see so much data out there about every single high school—test scores, graduation rates—but there's very little information around who your peers are going to be," de la Torre said.
Sartain agreed. "It was really surprising to me to see that students who were very high achieving could end up with very different outcomes, even going to a better school. It's just not something that's talked about a lot in the world of school choice.
"There's a lot of jargon around finding a good fit for your child," she continued. "This paper pushes the conversation around what data should be available to parents and what does 'good fit' really mean."
Vol. 35, Issue 24, Pages 1,12