Agent Skills
› brycewang-stanford/Awesome-Journal-Skills
› jedpsych-rebuttal
jedpsych-rebuttal
GitHub用于撰写教育心理学期刊R&R修改回复信。针对审稿人关于方法严谨性、透明度或机制检验的要求，结构化生成逐点回复，确保符合格式与匿名要求，不伪造结果。
Trigger Scenarios
收到R&R修改通知并准备回复
审稿人要求进行多层模型重分析、功效检验或增加透明度披露
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill jedpsych-rebuttal -g -y
SKILL.md
Frontmatter
{
"name": "jedpsych-rebuttal",
"description": "Use when writing the response to a Journal of Educational Psychology revise-and-resubmit. JEP R&Rs often demand added methodological rigor (multilevel modeling, cluster-level power, mechanism tests), fuller JARS disclosure, or stronger transparency, and the response must address every point while keeping the paper within format and masked. Structures the response letter; it does not fabricate new results."
}
R&R Rebuttal (jedpsych-rebuttal)
A Journal of Educational Psychology R&R (usually a major revision) typically asks for more methodological rigor (model the nesting, power at the cluster level, test the mechanism), fuller JARS disclosure, or stronger transparency — and all of it must fit back inside the format and stay masked. The response letter must convert every reviewer and reassure the handling editor on rigor and educational relevance, while keeping the manuscript within the word budget.
When to trigger
- An R&R arrived and you are planning the revision + response letter
- Reviewers requested multilevel reanalysis, added power, mechanism tests, disclosure, or open-science changes
- A requested analysis would change the claim
- Writing the cover note to the editor
Strategy
- Read the editor's letter as the rubric. Solve the decisive points first; the handling editor adjudicates among reviewers and synthesizes the required changes.
- Point-by-point, every comment. Quote each comment, then respond; never skip one.
- Strengthen rigor, don't just defend. Many requests (refit a multilevel model, re-power at the cluster level, add the mediation test, report effect sizes with CIs, fuller JARS disclosure, better data/code sharing) genuinely strengthen the paper — do them and say where. Update the Transparency and Openness subsection if deposits or preregistration status changed.
- Manage the word budget. New robustness, full measurement models, and extra specifications usually belong in online supplemental material; summarize briefly and point to the supplement so the body stays under the limit.
- Concede or rebut with evidence. Did what was asked (cite the location), or push back respectfully with a reason — don't add an analysis that quietly undercuts the claim without saying so.
- Keep the revision masked. No author names, grant numbers, site/IRB names, or first-person self-
citations in the revised manuscript or links; keep shared scripts/data in sync (see
jedpsych-open-science-and-transparency).
Response-letter format
For each reviewer comment:
> [Quoted reviewer comment]
Response: [What we did / why we respectfully disagree].
Change: [Main text section, supplement section, or table/figure number].
Open with a short summary of the main changes to the editor; group by reviewer; end each entry with the location (note when content went to the supplement to protect the word budget).
Worked micro-example (illustrative response entries)
For the cluster-randomized reading trial, an R&R asked for multilevel reanalysis, the mechanism, and fuller disclosure.
> R2: The analysis treats students as independent; with classroom
> randomization, the SEs are too small.
Response: Agreed. We refit the primary analysis as a two-level model
(students within classrooms) with random classroom intercepts and a pretest
covariate. The effect remains educationally meaningful (g = 0.23, 95% CI
[0.06, 0.40]; ICC = 0.14). Cluster-level power is now justified in Method.
Change: Method (power), Results (two-level model, Table 1); robustness → Supplement S2.
> R1: You claim a strategy-instruction mechanism but never test it.
Response: We added the preregistered multilevel mediation: gains in
metacognitive monitoring mediate ~40% of the effect (indirect 95% CI
[0.02, 0.13]).
Change: Results (mediation, Figure 1); analysis code updated in the deposit.
> R3: Attrition handling is unclear.
Response: We now report attrition by arm and re-estimate under FIML; the
effect is unchanged (g = 0.21).
Change: Method (attrition), Results (robustness); JARS flow reported.
R&R triage — where each request lands
| Reviewer ask | Default home (protects the word budget) | Note |
|---|---|---|
| Refit multilevel / re-power at cluster level | Method + Results (it is the contribution) | report ICC; update power justification |
| Mechanism (mediation/moderation) test | Results | preregistered if possible; update deposit |
| Added robustness / specification grid | online supplemental material | summarize in one main-text sentence |
| Fuller disclosure (exclusions/attrition/measures) | Method + Transparency subsection | a table, not prose, to save words |
| "Soften the claim" | Discussion | scale wording to the CI and the setting studied |
| Effect-size/CI + educational interpretation | tables and inline | route exhibits to jedpsych-tables-figures |
Recurring R&R pushback and the venue fix
- "Clustering ignored" → refit a multilevel/random-effects model, report the ICC, re-justify cluster-level power; never argue the single-level p-value. "Mechanism asserted, not tested" → add the preregistered mediation/moderation and interpret it.
- "You added an analysis that weakens the effect" → disclose it, interpret it, adjust the claim; concealment is fatal here. "Revision busts the format" → push new robustness to the supplement, cite it.
Anti-patterns
- Ignoring or merging away a comment without a visible response
- Cramming new robustness into the body and busting the word limit
- Capitulating to a change that breaks the paper's logic
- Adding analyses that contradict the original claim without acknowledgment
- Letting shared data/scripts/Transparency subsection drift out of sync, or reintroducing identifying information into the masked revision
Output format
【Editor's decisive points】addressed first? [list]
【Coverage】every reviewer comment answered? [Y/N]
【Rigor strengthened】nesting/power/mechanism/disclosure improved? [Y/N]
【Word budget】new material in supplement; body ≤ 12,000? [Y/N]
【Open-science updated】scripts/data/Transparency subsection in sync? [Y/N]
【Masked】no identity reintroduced? [Y/N]
【Next】resubmit via Editorial Manager
Supplementary resources
../../resources/official-source-map.md— masked review, JARS, transparency, and format policy
Version History
- 1839142 Current 2026-07-05 13:36
