On the remote learning ‘science’ (editorial)

Did remote learning contribute to “learning loss” during the pandemic? Some authors, including me, think so, but it’s not quite that simple.

While the harmful effects of remote learning have been well documented, even before the pandemic, the blame for learning loss must be shared between remote learning and several other factors, including the increase in stress on students, teachers, and families during a global pandemic.

The Washington Post published an article two days ago entitled “The science on remote schooling is now clear. Here’s who it hurt most,” which reports that, while most students lost ground during the pandemic, students living in poverty lost much more ground than those in more affluent communities.

The article generally uses hyperbolic language to describe the drops in scores on reading and math tests:

The article lays the blame for the score drops seen in reading and math on remote learning, citing a growing body of research, implying that the author, Laura Meckler, has concluded that those research papers reached the same conclusion she did for her article.

The first such study cited comes from Dan Goldhaber et al at Harvard University. Using data from reading and math scores on tests administered in several states, a drop in scores occurred between the spring of 2019, a year before the pandemic, and the spring of 2021, a year into it.

“While we have nothing to add regarding the public health benefits, it seems that the shifts to remote or hybrid instruction during 2020-21 had profound consequences for student achievement,” the study concluded. “In districts that went remote, achievement growth was lower for all subgroups, but especially for students attending high-poverty schools. In areas that remained in person, there were still modest losses in achievement, but there was no widening of gaps between high and low-poverty schools in math (and less widening in reading).”

The study’s conclusions clearly describe the positive correlation shown in the study’s data between remote learning and achievement loss, as measured by standardized tests in reading and math. While the achievement loss happened more in schools that went all-remote than in schools that brought students into classrooms, the study doesn’t say remote learning “caused” the loss in achievement.

“It is possible,” the authors write, and I would add entirely possible or even likely, “that the relationships we have observed are not entirely causal, that family stress in the districts that remained remote both caused the decline in achievement and drove school officials to keep school buildings closed.”

That’s a recognition that this may be a spurious correlation. In other words, family stress may be the root cause, not remote learning. That stress, probably more significant for families in poverty than more affluent families, is the cause of both correlated factors: achievement loss and a decision to keep school buildings closed.

Reading scores in Ill. show a drop in the % of students in the state who are “proficient”

We will never know because schools and teachers need to keep themselves busy recovering from the achievement losses, which are well documented, as shown above for the Illinois Assessment of Readiness test in reading in grades 3 through 8.

It’s too much to ask schools to study whether remote learning is to blame for achievement loss when they need to focus on recovery efforts. We don’t have time to conduct controlled studies to find out the root cause of achievement loss when we have lots of ground to make up.

The scientific leap from correlation to blame, absent those more controlled studies, is either invalid or superfluous (we have already documented the adverse effects of remote learning on these pages). We can’t just blame remote learning for this much achievement loss apart from other factors that increased or decreased as the global pandemic disrupted life in so many ways.