CSCW Author Response

There is no rebuttal box at CSCW. The venue's response instrument is the revision plus response letter of a journal: you get weeks or months, you may change anything in the manuscript, and the people reading the result are the reviewers who wrote the critiques. That combination — real time, real changes, remembered context — makes the response letter the highest-leverage document of the whole pipeline.

Triage: sort requests before touching the manuscript

Extract every distinct request from every review into a ledger, then classify:

Class	Definition	Response posture
Do	Legitimate, feasible in the window	Do it fully; show where
Do differently	Legitimate concern, wrong prescription	Fix the underlying concern; explain the substitution explicitly
Decline	Would harm the paper or exceed the study's license	Decline once, with reasons — never silently
Cannot	Requires data/access that no longer exists	Say so plainly; offer the honest partial (new limitation text, added analysis of existing data)

Two CSCW-specific triage realities:

• Cross-tradition requests are common. A quantitative reviewer may ask an interpretive study for reliability statistics; a qualitative reviewer may ask a trace study for member voice. Handle these as "do differently": meet the underlying worry (analytic credibility; human grounding) within your tradition's own rigor, and teach the standard you are meeting in one respectful sentence.
• New-data requests need early decisions. More interviews or another deployment wave may be impossible mid-revision. Decide in week one whether to collect, substitute, or descope a claim — this decision gates everything else.

Running the revision as a project

• Assign each ledger row an owner and a due date inside the window stated in the decision letter (confirm the date; under rolling it is set per paper).
• Sequence: analysis-changing work first, prose second, response letter third — a letter written before the changes stabilize will misdescribe them.
• Freeze a diffable copy of the round-one manuscript; you will need honest before/after references, and reviewers may check.

The response letter

Structure that survives a skeptical same-reviewer read:

Dear Editor and Reviewers,

Summary of changes (½ page max):
  the 3-5 revisions that changed the paper's substance, plainly stated.

Per-reviewer, per-point:
  R1.1  [quote or faithfully paraphrase the request]
        Response: what we did / why we did something else / why we declined.
        Location: §4.2, ¶3; Appendix C, Table 7.
  R1.2  ...
  R2.1  ...

Changes not requested by any reviewer:
  list them — silent edits discovered later cost trust.

Letter discipline:

• Answer the request that was made, then stop. Padding a response with re-argued contributions signals evasion to someone who has read it all before.
• Point to locations, not to effort. "We substantially revised Section 4" is unverifiable; "§4.2 now reports the attrition counts (new Table 3)" is.
• Keep the tone collegial and unwounded. The reviewer rereads your paper with this letter's tone in their ear.
• Preserve anonymity — the revision round is still anonymized: no names in the letter, no identifying artifacts in newly added materials.
• Concede real hits explicitly. "The reviewer is right that our round-one framing overclaimed; we have rescoped the claim to the six communities studied" converts a weakness into demonstrated judgment.

Revise-for-External-Review responses

The internal-round variant is smaller but the same shape: a short letter, the specific repairs, done fast. Its purpose is to make the paper worth reviewers' time — fix exactly what the screeners flagged and resist the urge to renovate.

Exit check before resubmission

[Ledger]    every request answered exactly once (no orphans, no repeats)? y/n
[Locations] every claimed change verified at its cited section/page? y/n
[Diff]      unrequested changes listed? y/n
[Tone]      letter read aloud by a co-author not involved in drafting it? y/n
[Anonymity] new text, tables, and supplements swept? y/n
[Deadline]  resubmission date and format confirmed against the decision letter? y/n

In the last verified cycle roughly four of five R&R papers that resubmitted were accepted (152 of 196; checked 2026-07-08) — the loop rewards teams that treat it as the venue's main event rather than an appeal.