Written Communication Certificate · Module 2 of 3
Feedback and Assessment in Writing
How feedback works, what the research says about what students actually do with it, and practical strategies — minimal marking, screencast feedback, and rubric design — that reduce faculty labor while increasing student learning. Estimated time: 4–6 hours over two weeks.
Rethinking Response: From Correction to Conversation
The dominant model of written feedback in higher education is correction — find what is wrong and mark it. This model is deeply familiar because it is how most faculty received feedback on their own writing. It is also, research suggests, largely ineffective for developing writers.
Sommers' foundational 1982 study found that teacher comments typically displaced students' own intentions, treating student writing as if it were a final draft to be corrected rather than a developing text. Students who received correction-focused feedback tended to make surface-level changes without engaging with substantive revision. Feedback that asks questions, identifies patterns rather than enumerating errors, and responds to the writer's meaning — rather than marking departures from correctness — produces significantly more substantive revision and better long-term development.
The shift from correction to conversation requires asking different questions. Instead of "what is wrong here?" the question becomes: "what is this writer trying to do, what is getting in the way, and what are the two or three most important things to address right now?" This reframing — from judge to coach — is the conceptual foundation of the module.
The feedback paradox
More feedback is not more helpful. Research consistently shows that comprehensive error-marked papers overwhelm students and produce less revision, not more. Students need to understand what they should prioritize — and a paper marked everywhere signals that everything is equally important, which is cognitively overwhelming and practically useless.
The Case for Minimal Marking
Richard Haswell's "Minimal Marking" (1983) offered a practical solution to a real problem: how do faculty provide useful feedback without spending unsustainable time on it, and without overwhelming students with correction? Haswell's approach — check-marking unedited surface errors for students to find and correct themselves — has since been developed into a broader philosophy of minimal response.
The underlying principle is that editing should be the student's job, not the instructor's. When faculty correct errors for students, they remove the cognitive work that would actually help students learn. A check in the margin says "there is an error here" and requires students to identify, analyze, and correct it — which produces more durable learning than reading a correction someone else has made.
Minimal marking in practice means: responding only to substantive issues (argument, evidence, structure, development) in your written comments; using check marks or brief codes for surface errors; limiting comments to the two or three most important issues per draft; and reserving comprehensive response for revision rather than first drafts. Faculty who adopt minimal marking consistently report both less time spent on feedback and better student revision between drafts.
Implementation note
Minimal marking requires transparency. Explain to students why you are not correcting every error — some will initially read minimal marks as insufficient attention to their work. A brief explanation ("I focus my written comments on the three issues that will most improve this draft") goes a long way toward student buy-in.
Screencast and Audio Feedback
Screencast feedback — recording yourself reading and responding to student work while your cursor tracks through the document — is one of the most significant developments in writing feedback research of the last two decades. Anson's work (2015) and Davis and Vilhotti's research (2019) both document substantial differences in how students receive and use audio-visual feedback compared to written comments.
The mechanism appears to be that hearing a voice contextualizes feedback in ways that written comments don't. Students report understanding what instructors mean more clearly, feeling that instructors see them as individuals rather than papers, and feeling less defensive about critical feedback. Davis and Vilhotti found that students were substantially more likely to revise when they received screencast feedback, even when the content of the feedback was identical to written comments.
Tools include ScreenPal (formerly Screencast-O-Matic, free up to 15 minutes), Loom (free for educators), and QuickTime on Mac. The workflow is straightforward: open the student's paper, open the recording tool, record yourself reading through and commenting aloud. The recording is then shared as a link in comments or via email. Most faculty report that once they are comfortable with the format, screencast feedback takes the same time as written comments or less — because thinking aloud is faster than crafting careful written prose.
Rubrics: Tool or Trap?
Rubrics are among the most overused and underdesigned tools in higher education assessment. The promise of rubrics — transparent criteria, consistent grading, reduced student anxiety — is real. The failure mode — rubrics that fragment writing into disconnected criteria and reward formulaic compliance — is also real and common.
The problem with most rubrics is that they treat writing as a set of separable components (thesis, evidence, organization, mechanics) that each warrant a distinct score. This structure implies that writing is modular and that deficiency in one area is independent of others. Any experienced writing teacher knows this is false: a weak thesis is often inseparable from weak evidence selection and underdeveloped structure. Holistic rubrics — which describe the overall qualities of papers at different levels — are often more accurate and more useful than analytic rubrics, even though they feel less precise.
When analytic rubrics are appropriate, design principles matter. Criteria should reflect the genuine goals of the assignment, not generic writing qualities. Descriptors at each level should describe what the writing does, not what it fails to do ("the argument presents clear evidence for each claim" rather than "lacks sufficient evidence"). And rubrics should be developed with writing centers or writing specialists, shared with students before they write, and revised based on what you observe actually happening in student work.
When rubrics help and when they don't
Rubrics are most useful for assignments with multiple, genuinely independent criteria — complex projects, portfolios, oral presentations with distinct evaluable components. For a focused essay or a lab report with a clear genre, descriptive feedback is often more useful and more honest about the nature of the evaluation.
Peer Review as Pedagogy
Peer review is frequently dismissed as a strategy of uncertain value — students don't know what good writing looks like, so why ask them to evaluate each other's drafts? This dismissal misses the point. The primary beneficiary of peer review is not the writer receiving feedback; it is the reviewer. Reading multiple drafts at a similar stage of development, applying the criteria of an assignment to someone else's work, and articulating what is and isn't working — these are high-value metacognitive tasks that develop the reviewer's own understanding of what good writing looks like.
Effective peer review requires design. Asking students to "give feedback" on a peer's paper without structure produces vague, unhelpful responses ("This is good! Maybe add more detail?"). Effective peer review protocols give reviewers specific tasks: "Summarize the argument in two sentences," "Identify the three most important claims the paper makes and assess whether each is sufficiently supported," "Mark one sentence per page that seems unclear and explain what confuses you." Structured reviewer tasks produce substantive, useful feedback.
The logistics of peer review also matter. Reviewing a peer's draft before finalizing your own means you see a range of approaches to the same problem. Reviewing two or three drafts rather than one gives reviewers a basis for comparison. Reviewing the same draft twice, before and after the writer's revision, helps reviewers understand how feedback is (and isn't) used.
Activity 1
Feedback Comparison Analysis
Below are three versions of written feedback on the same student paragraph — one substantive and process-focused, one correction-focused, and one that combines surface correction with buried substantive concern. Read all three, then respond to the analysis prompts below.
Student paragraph (anonymized)
Version A — Correction-focused
"'Has been shown' is passive voice — rewrite. 'Several studies' — which studies? Be specific. The Oxford citation is incomplete. 'Regularly' is vague — how often? 'This suggests' is a weak transition. The conclusion overstates the evidence significantly."
Version B — Conversation-focused
"You're making a significant causal claim here — that chocolate could treat depression — but your evidence only shows a correlation between chocolate consumption and mood. What's your argument for moving from correlation to treatment? That's the gap I'd want you to address in revision. Also: when you cite 'one study,' readers will want to know enough to find it themselves."
Version C — Mixed
"Passive voice throughout — avoid. The studies you cite need proper citations. Overall the logic of this paragraph is a bit shaky — you move from correlation to causation without justification. Fix the grammar issues first, then think about whether your conclusion is actually supported by your evidence."
Contribute to the repository
If you wrote a revised version of your own feedback as part of this activity, that's a resource others would find valuable — especially paired with your analysis of why you changed it.
Activity 2
Screencast Feedback Practice
The goal of this activity is to produce one screencast feedback recording on a real piece of student writing from your course. You will then reflect on the experience.
Before recording: set up a screencast tool (Loom, ScreenPal, or QuickTime on Mac). Open the student's document in a format you can scroll through on screen. You don't need to prepare a script — the value of screencast feedback is thinking aloud in real time.
- Record a 3–5 minute screencast response to one student draft. Track through the paper with your cursor as you read. Respond to substance first — what is the student trying to do, what is working, what is the most important issue to address. You don't need to address everything.
- After recording, watch the first 60 seconds of your recording. Notice: What do you lead with? Do you start with strengths or problems? Does your tone match how you'd talk to a student in office hours?
- Share the recording with the student. (A Loom link in a comment, or emailed, both work.)
- Follow up with the student — either in office hours or a brief email — asking one question: what was most useful about the screencast, and was anything confusing?
Contribute to the repository
Your protocol for integrating screencast feedback — which assignments, what length, how you share the link — would be a useful model for colleagues in your discipline. You don't need to share the recording itself.
Activity 3
Rubric Audit and Redesign Workshop
Select one rubric you currently use (or have used) for a writing assignment. Audit it against the principles in Section 4, then redesign one section of it.
- Do the criteria reflect the genuine goals of the assignment, or generic writing quality?
- Do descriptors describe what the writing does, or what it fails to do?
- Are any criteria genuinely separable, or are some deeply intertwined?
- Would a student reading this rubric before writing know what to do? Or only know how they will be judged after the fact?
Contribute to the repository
Discipline-specific rubrics are among the most frequently requested resources in the repository. A rubric from your discipline — especially one you've tested and refined — is directly useful to colleagues teaching similar assignments.
Key readings and resources
Three marked ★ are the most essential for the module's frameworks.
- ★ Sommers, N. (1982). Responding to student writing. College Composition and Communication, 33(2), 148–156.
- ★ Anson, C. M. (2015). Habituated practice can lead to bad teaching. College Composition and Communication, 66(4), 615–625.
- ★ Davis, M., & Vilhotti, J. (2019). Screencast feedback and student perceptions of instructor presence. Computers and Composition, 53, 1–14.
- Haswell, R. (1983). Minimal marking. College English, 45(6), 600–604.
- Straub, R. (1997). Students' reactions to teacher comments: An exploratory study. Research in the Teaching of English, 31(1), 91–119.
- Elbow, P., & Sorcinelli, M. D. (2006). How to enhance learning by using high-stakes and low-stakes writing. In W. J. McKeachie & M. Svinicki (Eds.), McKeachie's teaching tips (12th ed.). Houghton Mifflin.
- Nilson, L. B. (2015). Specifications grading: Restoring rigor, motivating students, and saving faculty time. Stylus.
Complete Module 2
When you have completed the three activities and are ready to move on, submit your responses below. Your work is saved locally in your browser — this submission sends a copy to CWSP.
You will receive a confirmation email. Module 3 will be available immediately after submission.