A new study by Georgia State University finds that many students didn’t learn as much as they might have had the pandemic not occurred. But the results vary widely based on family wealth, race and other demographic factors, and also by grade level and subject.

The study, one of the first comprehensive local looks at the effect of the pandemic on learning, found what most observers have feared: the schooling disruptions over the past year caused many students to learn less than they would have, though some findings surprised one of the lead researchers.

“I would have thought students with disabilities would have been impacted more,” said economics professor Tim Sass. He also did not expect the extreme variation by grade level that his team found.

Middle school students did worse in math and reading than younger students.

The research by the Georgia Policy Lab in the Andrew Young School of Policy Studies released Thursday used scores from proprietary “formative” tests used by three unidentified large metro Atlanta school districts. State standardized tests were canceled last spring and not taken by all students this school year. GSU says their study is one of two independent, multi-district studies to estimate the pandemic’s impact on student achievement in the United States.

While many talk of learning “loss,” GSU is describing their findings as a slowing of anticipated “achievement growth,” based on what would have been projected in a normal year from test score trends preceding the pandemic.

In some cases, that deceleration equaled what would have been learned over three-quarters of a normal school year. It was typically worse for students who remained home and online, but even those attending school in-person generally fell behind where they should have been. The most damage occurred this school year rather than last spring.

“Across most measures, students are three to six months or more behind where they would have been had the pandemic not occurred,” the researchers wrote.

For instance, in one of the school districts, eighth-grade students on average learned the equivalent of 6.9 months less than expected in math. In another, seventh-grade students learned an average of 7.5 months less in reading than projections showed they should have.

Reading was down in varying amounts in all of the measured grade levels — four through eight. Math was more inconsistent, though. For instance, one district had test scores showing its eighth-grade students learned a month more than projected for math.

Poverty played a mostly predictable role, generally correlating with poorer performance. However, even here the results varied. In one district, eighth-grade students who qualified for free or reduced-price meals were behind by 1.6 months in reading while peers who didn’t qualify, presumably because their family income was too high, were 2.4 months behind.

Sass said the findings may be instructive statewide as school districts plot their recovery from the pandemic. GSU is recommending targeted remediation using federal stimulus funds — tutoring, longer school days, summer school — for those who need it, rather than broad brush approaches affecting all students.

Clayton County Superintendent Morcease Beasley was among those watching the online presentation.

“Tough year for all but really tough for those already negatively impacted by inequity in our society,” he posted in the “chat” rail. “The inequity was perpetuated through laws, policies, and discriminatory practices.” Schools should be “intentional” about eradicating those inequities, he wrote.

Ryan Moore, a Fulton County Schools administrator, was among the presenters. He said his district was already following GSU’s recommendations, saying of Sass’s proposals, “It’s almost like you’re reading off the Fulton County recovery plan.”