Feedback and the Kritik cycle

Peer review that is itself marked

Twice in the term you submit a piece of writing to Kritik, a calibrated peer-review platform, and then grade classmates' pieces against the same rubric. The 10 % is not for the piece — it is for how well you grade and how useful your feedback is. That makes Kritik the one place in the course where rubric literacy is marked directly.

A Kritik activity runs in three stages (the platform's own labels, read at time of writing, September 2026):

1. Create — you submit your piece by the deadline.
2. Evaluate — you grade a small number of anonymous peers' pieces, rubric line by rubric line, and write a comment on each.
3. Feedback — you rate the evaluations you received: were they accurate, and did the comments help?

Calibration is separate. It is its own activity type, not a stage of every assignment: the instructor pre-grades a few past submissions, you grade the same ones, and the gap between your scores and the instructor's measures how trustworthy your grading is. A well-calibrated grader's scores carry more weight when peers' marks are computed. So the honest answer to "what are the stages?" is three verbs; calibration is the thing that happens beside them.

Worked example — one cycle, two comments

You submit a practice email (create). In evaluate you receive two peers' emails and the ten-line email rubric. For the first email you score each line and write:

> Your email is good but could be a bit stronger. Nice tone!

For the second you write:

> Line 2 (ask in sentence one): your first sentence is "Hope your week is going well" and the ask is in paragraph two — moving "Could you approve the revised budget by Friday?" to the top would earn this line. Line 5 (closing with a deadline): earned — "by Friday 4 October" is there. The subject line (five words, specific) works well.

In the feedback stage, the first writer rates your comment low and the second rates it high, and the platform is right to reward the difference. The first comment names no rubric line and no place in the text, so the writer cannot act on it. The second names the line, cites the sentence, says what would earn the line, keeps the tone respectful, and notes one thing that worked — which is exactly the feedback rubric this course uses (R-feedback, introduced properly in module 11): names the rubric line · cites the place in the text · says what would earn the line · tone respectful · notes one thing that worked.

Kinds of feedback

Feedback comes in two pairs that are worth telling apart, because the course asks you to seek all four:

• Direct feedback comes from the receiver themselves (the TA's marked rubric; a peer's Kritik comment). Indirect feedback comes from somewhere else — a classmate mentions that the group found your summary confusing.
• External feedback comes from other people. Internal feedback is your own review before anything is sent: reading the draft against the rubric, the monitor step of the metacognition cycle.

The TA's rubric is external, direct. Your own pre-submission audit is internal. A Kritik peer comment is external and direct; a rumour that your group's memo was "hard to follow" is external and indirect — less reliable, but not worthless.

Receiving feedback you disagree with

Two failure modes. Feedback that is praise-only ("great job") or fault-only ("this is wrong") is not constructive — it gives the writer nothing to do. And when you receive feedback you think is wrong, the professional move is to clarify, not defend: ask which line the grader had in mind and what would have earned it. Defending the draft ends the exchange; clarifying it produces the change you can put into the next plan.