10/24/2026
A writing assessment can give us a score. But how much can that number really tell us about a student?
Consider everything that goes into a student's performance: what they've learned, certainly, but also their motivation, the timing of an assignment, their personal circumstances, and the particular conditions of the classroom. Some of those influences are easy to recognize. Others are much harder to measure.
In “Empirical Writing Assessment and the Identity Crisis in Science,” Michael J. Zerbe asks what happens when we bring that complexity into our understanding of assessment.
As Zerbe shared with us, the essay began with a question: Why wasn't empirical writing assessment more meaningfully engaging with contemporary debates about science, especially when so much of our assessment work relies on the scientific method?
Drawing on debates about complex networks, replicability, and Big Data, he considers what those conversations might teach us about the evidence we collect. More data can reveal patterns, but those patterns don't necessarily explain what caused them—or whether we'd find the same results in another classroom.
His argument isn't that empirical assessment has no value. Rather, he invites us to treat its findings with greater rhetorical scrutiny, consider the contexts that shape them, and place them in conversation with the expertise of the people doing the teaching.
When we asked what he hoped readers would take away, his advice was simple: “Approach empirical writing assessment methodologies and results with a healthy degree of skepticism.”
After all, the numbers don't interpret themselves.
What's one thing an assessment score might miss about your students?
Find Zerbe's article in Composition Studies 54.1. Check your campus library for access, or explore the issue and subscription options through the link in our bio.