L39 — Peer Review

L39 — Peer Review#

Your complete draft is due at the start of class, and the period is peer review. You calibrate on one sample paragraph, then review one classmate's paper against the same rubric that grades yours — every line banded, every comment specific. One classmate reviews yours, and that review drives your revision at L40.

Before class#

Submit your complete draft as a PDF through the course Teams team before class starts. The review is written in this period, so a draft that is not in Teams at the start cannot be reviewed. Late deliverables need coordination with the instructor at least 24 hours before the deadline; otherwise they receive a zero.

Today’s clock#

Clock	What happens
0–15 min	Drafts in through Teams, reviewers posted; calibrate on one sample paragraph, marked against the rubric
15–45 min	Your review, on the peer-review form
45–50 min	Review submitted through Teams and returned to its author

How drafts and reviews move#

Your draft goes to the course Teams team as one PDF. Every paper gets one reviewer: in the roster order the instructor posts, you review the cadet after you, and the cadet before you reviews you (the last reviews the first). You give one review and receive one.
Your reviewer reads the PDF and nothing else — your code is not exchanged.
Your review is written in class today, on the peer-review form (printed, or an annotated copy of the PDF), and submitted as a PDF through Teams before you leave. It is returned to its author for L40.
The author revises from that review at L40.

The peer-review form#

The form carries the Final Project rubric’s eight lines. For each line you circle one band and write one specific comment:

Exceeds — complete, correct and clearly argued, beyond what the task requires.
Meets — complete and correct.
Developing — present, but incomplete, unclear, or with errors.
Missing — not in the draft.

Two lines are about the review process rather than the paper: line 6 (Peer review given) is a self-check — you band your own review before you submit it — and line 7 (Response to reviewer) is not banded on a draft; there you name the one comment the author’s response table most needs to answer.

Then three more boxes:

The citation check: are all four blocks cited, and does one cited lesson you open support its claim?
The strongest element of the paper, and where it is.
The single change that would most improve the paper.

How your review is graded. Your review is the rubric’s Peer review given line — 5% of your own Final Project grade. It is graded on being specific (every comment points at a page, section, figure or equation), accurate (right about the models, the numbers and the rubric), and anchored to the rubric (each comment sits on its line, and the band you circle matches what you wrote). “Looks good” and “add more detail” earn nothing.

What a good review looks like#

A comment the author can act on names where, says what is wrong or strong, ties it to one rubric line with a band, and says what would fix it.

	Weak	Good
Comment	“The results section is confusing and could use more detail. Nice figures overall!”	“Table II, p. 2: the unsupported row repeats the aspect-blind row’s promised \(P_S\), so the route labeled aware used nose-on \(\sigma\). Route results and the aspect-blind comparison — Developing. Re-plan it with the aspect-aware edge risks Sec. III states.”
Specific	No page, section or figure	Table, page and row named
Accurate	Nothing to check	Names the model error exactly
Anchored	No rubric line, no band	One line, one band, one fix

Comment on the paper; do not rewrite it. And check before you “correct” a number: different routes give different numbers, so the question is whether the author’s number follows from the author’s stated method.

The citation check#

The paper must cite at least one course lesson from each block, each for a specific claim; external sources are optional.

Circle the blocks cited on the form: is there at least one lesson from each of Blocks 1–4?
Pick one cited lesson that carries a claim the paper depends on.
Open it from the URL in its reference: does the lesson support the claim it is cited for, where it is cited?
If the paper cites an external source, is it credible and publicly released, with a one-line vetting in Appendix A?

A block with no lesson, or a citation that does not support its claim, is a finding: write it down.

When your review comes back#

At L40 you revise from the review of your paper and write the response-to-reviewer table (ResponseToReviewers.tex, in the template): every comment that asks for a change gets a row — form lines 1–5 and 8, the citation check and the single change, in the form’s order; lines 6 and 7 and the strongest element need none. If you changed something, say what and where in the final paper; if you did not, say why, with your evidence. Disagreeing with your reviewer is fine — ignoring a comment is not. The table is worth 5 points and is due with the final paper at the start of L41. See L40 — Revision Work Day.