It’s not all about the test scores, stupid.
That sums up a new University of Chicago study, a groundbreaking analysis of middle-school student performance that lays out which measures best predict success in high school and college.
What matters most for later academic success are middle-school grades and attendance, far more than test scores and demographic factors (race, poverty and the like), concluded the study of Chicago Public Schools fifth- through eleventh-graders. Standardized test scores are not the best predictors of academic success, as our test-crazed world might have us believe.
The real-world implications are clear: the Chicago Public Schools should continue to scale back its intense focus on standardized tests and turn to what matters — boosting middle-school grades and attendance.
The researchers found that attendance and overall grade-point average in middle school were the strongest predictors of actual school performance in ninth grade and 11th grade, both of which strongly predict high-school graduation rates and college success.
This new finding builds on a similar, well-established finding for high schools: grades are by far the most important predictor of getting into college and eventually graduating, more so than ACT or SAT scores or high school coursework.
“Test scores are very good at predicting future test scores but not as strongly predictive of other outcomes we care about, like whether students will struggle or succeed in high school coursework or graduate from college,” Elaine Allensworth, director of the university’s Consortium on Chicago School Research and lead author of the study, said in a statement.
Good grades reflect mastery of skills valuable in college and in life in general, such as broad knowledge (not just reading and math), writing and capacity for sustained effort. Standardized tests, in contrast, hone in on a far more narrow band of skills. They have value, but too tight a focus on test prep is counterproductive.
“It actually discourages teachers from spending time on things that are more important for kids,” Allensworth said in an interview.
The researchers also found far less variability in what constituted an A or B grade at schools across the city than one might suspect. There is variability, as much as a half grade point at the extreme, but it’s not enough to undermine the predictive value of grades.
Other key take-aways from the study:
◆ These findings allow schools to identify kids as young as fifth grade at risk of failing in high school based on their grades and attendance records, and to target intervention to boost those two areas, which are more malleable than test scores.
The researchers found, for example, that students’ probability of being on track in the ninth grade goes from 66 to 93 percent, depending on whether their attendance declines (from 97 to 93 percent) or improves (from 97 to 99 percent) in the middle school grades. Students with attendance below 90 percent in middle school are at high risk of not graduating from high school.
◆ The researchers also found that some middle and high schools do far better intervening in these key areas than others. Among middle-school students that saw improvements in attendance, grades and test scores, about half of the differences could be attributed to the school they attended, they found. Schools can have considerable influence in particular over attendance.
CPS in the last year has made improving elementary attendance a priority. Chronic absenteeism, which has been unacceptable high for years, dropped slightly last year, though it’s still early. The school system also has improved high-school attendance rates and significantly improved freshmen pass rates, which is a key predictor of graduation rates.
The freshman on-track rates went up after CPS developed data systems to closely track freshmen and added supports to redirect wayward students. Using the U. of C. report as a guide, CPS should quickly set up a similar data system to track middle school attendance and grades and give schools the time and support they need to make good use of the data. That’s in the early stages, we’ve been told. We urge CPS to make it a top priority.