What is the right distance for spacing students in classrooms?

Infection and Spread School

A: There is no magic distance, at 3 feet or 6 feet, that the coronavirus, or any virus for that matter, respects.

A recent study has generated lots of buzz with its conclusion that 3 feet of physical distance is just as good as 6 feet of space between students (link below). But the study’s limitations make it difficult to draw strong conclusions about the head-to-head comparison of 3 or 6 ft of space.

Let’s pull out our trusty three C’s to do some data diligence: Comparison, Chance, and Context. We like to totally nerd out with the Three C’s – so apologies for the lengthy post! Feel free to skim the emoji bullets for the TL;DR 🤓

📊 COMPARISON: How credible is the comparison between study schools with a 3-ft policy versus those with a 6-ft policy? It’s not an apples-to-apples comparison.

Nerdy Girl Rating: 🛑CAUTION ADVISED🛑

The researchers compare COVID case rates in Massachusetts school districts providing in-person schooling with a 6-ft distancing requirement to those using a 3-ft distancing guideline. The case rates are based on student and school staff cases reported to the state. They find no (statistically detectable) differences in case rates based on the distancing rule and conclude that a guideline imposing 3 ft between students is sufficient for safe in-person schooling. It is difficult to isolate the effects of a distancing rule because the school districts that choose the 3-ft rule and those that choose the 6-ft rule may be different in other ways, including disease prevalence in the community and other non-pharmaceutical interventions (NPIs) layered in their transmission prevention strategies. Many of these differences simply can’t be measured and accounted for in the statistical analyses. For example, we may worry that the school districts that elected to have a 3-ft policy did so because of their knowledge of other disease mitigation measures in the community (likely leading to low out-of-school transmission) including behavioral compliance with public health guidance.

An additional wrinkle: the 3 versus 6 ft distinction is measured using self-reports (in school district plans), not what was actually happening in schools, which adds measurement error into the key variable of interest. Measurement error of this sort could hide any true differences in cases across the two types of school district. For our fellow Super Nerds, we’ve added a related reference below.

📊 CHANCE: How likely were the results to have arisen from the play of chance? The results have very large margins of error.

Nerdy Girl Rating: 🛑CAUTION ADVISED🛑

We want to know that any difference – or lack of difference! – observed by the researchers represents the truth, and not just random chance. While the research team has a reasonably large number of districts observed over many weeks to track weekly cases, their estimates cannot rule out slightly better COVID metrics in those places using a 6-ft rule. These data, therefore, do not provide a definitive take about the difference between schools with a 3-ft requirement and those with a 6-ft rule.

📊 CONTEXT: How well do the results translate beyond the research setting? This is a single-state study in Massachusetts. We can’t assume that what happens in Massachusetts schools generalizes well to Mississippi (or Minnesota, or Missouri, or Montana….)

Nerdy Girl rating: 🛑CAUTION ADVISED🛑

Context is really critical here as the study relies on data from Massachusetts districts with universal masking in a state where mask-wearing is high. These school districts are likely also employing other layered prevention strategies and we already know, from other studies on the role of schools in COVID spread, that low community prevalence is key to safe in-person schooling. We should use caution in extending these findings to thinking simply about the physical distancing rule in other places.

A final note:

There is no special magic to 3 ft or 6 ft. Think of it like a speed limit – it is not the case that you are suddenly safer when driving 65 miles per hour than you were at 66 miles per hour, but there is a cutoff that represents that at higher speeds, driving becomes more dangerous. Similarly, the key point with COVID is that distancing reduces spread, and trying to maximize distance ensures that transmission is more challenging for the virus.

It’s hard (!) to statistically isolate the impact of any one mitigation measure – which is why we don’t have the perfect study comparing the 3-ft distancing rule to the 6-feet rule in schools, holding everything else the same. Importantly, there are examples of schools operating safely in-person with little to no in-school transmission of COVID-19. In these places, the common denominators appear to be layered NPIs, including universal masking, and the context of low community spread.


Link to study, forthcoming in the journal Clinical Infectious Diseases

Earlier Dear Pandemic post on COVID Spread and In-Person Schooling

This morning’s Dear Pandemic post on CDC school guidance

Measurement error explainer

Link to original FB post