Students cannot produce what they cannot define. Most writing assignments give students a task and a deadline. What they rarely give is a working picture of quality specific enough to guide decisions during writing—not what the teacher will evaluate, but what the student should be building toward.
The gap shows up in a familiar place: the rubric handed out with the assignment. Rubrics describe good writing accurately: “Develops ideas with specific evidence.” “Demonstrates clear organizational structure.” “Maintains a consistent claim.” While these are correct descriptions of what strong writing does, they are not instructions for how to produce it.
The difference between a description and a target
A student reading “develops ideas with specific evidence” at the top of an assignment does not know what to write next. The phrase describes a quality of finished work in evaluator language. It says nothing about what developing an idea actually involves—what counts as specific, how much evidence is enough, or how evidence connects back to a claim. A rubric tells the student what the teacher will be looking for. It does not tell the student what decisions to make during drafting.
Dylan Wiliam, in Embedded Formative Assessment (2011), draws this distinction carefully. Learning intentions and success criteria only do instructional work when students can interpret and apply them—not just receive them. A success criterion on the board is not the same as a student understanding what it means for their own writing. The latter requires the criterion to be concrete enough to prompt a decision.
Grant Wiggins made a related argument in Educative Assessment (1998). Evaluative feedback—telling a student their organization was unclear—describes the outcome. Educative feedback tells the student what to do differently. The rubric, as typically deployed, operates on the evaluative end. It arrives at submission, describes what was or wasn’t there, and offers no mechanism for the student to act on it before the grade is final.
The model problem
Part of the reason students don’t know what good work looks like is that they rarely see enough of it during instruction. A single exemplar shown once at the start of a unit is not the same as regular, referenced exposure to what quality writing does at the sentence level, the paragraph level, and the argument level.
Wiggins argued that students need to encounter models the way practitioners do—repeatedly, comparatively, with attention to the specific decisions that distinguish stronger work from weaker work. Not “this essay earned a 4 because it has a clear thesis” but “notice how this writer returned to the central claim in paragraph three and sharpened it—here is what that decision looks like, and here is what it produced.” That kind of model analysis takes time. It is also not separable from instruction. It is instruction.
What changes when the criterion is specific
The fix is not a longer rubric. A rubric with twelve descriptors at four levels is not more useful than one with three—it is harder to use, and it does not solve the underlying problem, which is that evaluative language does not translate into compositional decisions—students still will not understand what to write to achieve the expectations in the rubric.
The fix is criteria that describe what a writer does, not just what a finished product looks like. “Your evidence supports your claim” is evaluative. “After each piece of evidence, write one sentence that connects it back to your central argument” is instructional. Both point at the same quality. One gives the student something to do with it.
The research on formative assessment—Paul Black and Dylan Wiliam’s 1998 synthesis “Inside the Black Box” most prominently—is consistent on this point: students improve when they understand what quality looks like and have a mechanism to close the gap between where they are and where they need to be. A rubric shared only at submission satisfies neither condition.
Students don’t produce better writing because they’ve been told to. They produce it because they understand, specifically enough to act on it, what better writing requires.
Further reading
Paul Black and Dylan Wiliam. “Inside the Black Box: Raising Standards Through Classroom Assessment,” Phi Delta Kappan, 1998. The foundational synthesis on formative assessment—establishes that feedback improves outcomes only when students can use it to adjust their work before it is finished.
Dylan Wiliam. Embedded Formative Assessment. Solution Tree, 2011. Practical extension of the Black and Wiliam research; the chapters on learning intentions and success criteria speak directly to the problem of criteria students cannot act on.
Grant Wiggins. Educative Assessment: Designing Assessments to Inform and Improve Student Performance. Jossey-Bass, 1998. Wiggins’ distinction between evaluative and educative feedback remains one of the clearest framings of why assessment design either teaches or merely judges.
W. James Popham. “What’s Wrong—and What’s Right—with Rubrics,” Educational Leadership, October 1997. Popham’s argument that rubrics often describe traits of quality without teaching students how to achieve them—a structural critique that holds up across the decades since.