Use peer review to improve the proposal · 利用同行评审改进提案
“Looks good” cannot repair a method mismatch
- A peer reads the materials proposal and writes, “Nice topic and neat layout.” The encouragement is welcome, but it misses a problem: one draft paragraph still says the team will ask whether participants like the labels, while the question concerns sorting accuracy.
- This original GAC033-aligned lesson treats peer review as a reasoned check of the proposal. It is not permission to run a study or a judgement of the author's character. Useful feedback identifies a specific issue and why it matters.
“看起来不错”无法修复方法不匹配
- 一位同伴阅读材料提案后写道，“不错的题目和整洁的布局”。鼓励值得欢迎，但它遗漏了一个问题：一份草稿段落仍表示团队将询问参与者是否喜欢标签，而问题却关乎分类准确性。
- 本原创GAC033对齐课程将同行评审视为对提案的理性检查。它不是进行研究许可或对作者品格的判断。有用的反馈会指出具体问题及其重要性。
Why is “nice topic and neat layout” insufficient as the only review? · 为什么“有趣的题目和整洁的布局”作为唯一评审是不够的？
Positive surface comments can miss a substantive mismatch. Encouragement and critical examination can coexist. · 表面的正面评论可能忽略实质性的不匹配。鼓励与批判性审视可以共存。
Review the argument before the appearance
- Read in this order: purpose and question, evidence, method, ethics, analysis, feasibility, then clarity and presentation. A well-written sentence can still promise a measure the procedure never collects. Check the connections between sections.
- Apply criteria consistently. Does the background support the question? Is the outcome measurable? Could another helper follow the procedure? Is participation voluntary? Are claims limited to the sample and task? Do not require the team to obtain your preferred result.
在外观之前审查论证
- 按以下顺序阅读：目的和问题、证据、方法、伦理、分析、可行性，然后是清晰度和呈现。写得很好的句子仍可能承诺程序从未收集过的测量。检查各部分之间的联系。
- 一致地应用标准。背景是否支持问题？结果是否可测量？另一位助手能否遵循程序？参与是否自愿？主张是否仅限于样本和任务？不要要求团队获得你偏好的结果。
Order the priority of these review tasks. · 安排这些评审任务的优先级顺序。
Substantive coherence and safeguards come before surface polish. · 实质连贯性和保障措施优先于表面润色。
Write a comment with a reason and a next step
- Use location → issue → consequence → suggestion. “Method paragraph 2 asks about preference, but the question concerns complete sorting success. This would leave the main question unanswered. Use the four-item task and keep preference, if collected at all, as a separate outcome.”
- Distinguish a necessary correction from an optional idea. Correcting the mismatch is necessary; adding a decorative colour theme is optional. If a detail is genuinely unclear, ask for clarification instead of claiming it is definitely absent from the whole project.
Specific feedback: “The consent script says sheets are anonymous, but the draft sheet asks for names. Remove the name field or explain and justify a different approved data plan; the current versions contradict each other.”
写出带有理由和下一步骤的评论
- 使用位置 → 问题 → 后果 → 建议。“方法段落2询问偏好，但问题关乎完全分类成功。这将导致主要问题未被回答。请使用四项任务，并将偏好（如果收集的话）作为单独的结果保留。”
- 区分必要修正与可选想法。修正不匹配是必要的；添加装饰性配色方案是可选的。如果某个细节确实不清楚，应请求澄清，而不是断言其完全不存在于整个项目中。
具体反馈：“同意脚本称表格匿名，但草稿表格却要求填写姓名。请移除姓名栏或解释并证明另一种已批准的数据计划合理；当前两个版本相互矛盾。”
Which TWO details make a feedback comment actionable? · 哪两项细节使反馈评论具有可操作性？
Location, issue and consequence identify a revision target. Personal attacks and preferred results are inappropriate criteria. · 位置、议题和后果标识出修改目标。人身攻击和期望结果是适当的评审标准。
Receive criticism without defending every sentence
- Pause and restate the concern before replying. “You noticed that our question and survey measure differ” shows understanding. Ask which part needs clarification and compare the comment with the actual draft, rather than assuming criticism means the project is worthless.
- You need not accept every suggestion. Reject one with a reason tied to scope or evidence: “We will not add a memory test because our question concerns immediate sorting and the extra task would change the approved procedure.” Respectful disagreement is compatible with useful review.
Peer approval does not replace teacher approval, participant information or source verification. A positive review can still overlook a problem. Keep the relevant checks in the action plan.
接受批评时不必为每一句话辩护
- 回复前暂停并重述担忧。“你注意到我们的提问和调查测量方式不同”表明你已理解。询问哪一部分需要澄清，并将评论与实际草稿对比，而不是假设批评意味着项目毫无价值。
- 你不必接受所有建议。因范围或证据原因拒绝其中一项：“我们不会增加记忆测试，因为我们的问题关注即时分类，额外任务会改变已批准的流程。”尊重的异议与有用的评审并不冲突。
同伴认可不能替代教师认可、参与者信息或来源验证。积极评审仍可能忽略问题。请在行动计划中保留相关核查步骤。
Every peer suggestion must be accepted even when it changes the project beyond its approved scope. · 即使建议改变了项目超出批准范围，也必须接受每一个同伴建议。
Consider suggestions and give reasons. A change outside scope may require rejection or further approval. · 考虑建议并给出理由。超范围的更改可能需要拒绝或进一步批准。
Make revisions traceable
- Keep a response log showing the comment, decision, change and reason. Mark accepted, partly accepted or not accepted; then identify the revised section. “Done” alone does not tell the reviewer whether the underlying issue was fixed.
- Recheck linked parts after a change. Removing a name field may affect data handling and withdrawal instructions. Changing the outcome requires updating the question and analysis. A revision that fixes one paragraph but leaves contradictions elsewhere is incomplete.
| Comment | Decision | Revision and reason |
|---|---|---|
| Preference measure does not answer accuracy question | Accept | Restore the fixed sorting task and success count |
| Add a second memory task | Do not accept | Outside the question and current scope |
| Name field conflicts with anonymous plan | Accept | Remove field and check information script |
使修订可追溯
- 保存回应日志，记录评论、决定、变更及理由。标记“已接受”、“部分接受”或“未接受”，然后指出修订的具体章节。仅写“已完成”无法让评审者知晓根本问题是否得到解决。
- 变更后重新检查关联部分。移除姓名栏可能影响数据处理和退出说明。更改结果需更新提问和分析。若仅修正一段文字而其他地方仍存在矛盾，则修改不完整。
| 评论 | 决定 | 修订内容及理由 |
|---|---|---|
| 偏好测量未能回答准确性问题 | 接受 | 恢复固定的分类任务和成功计数 |
| 增加第二个记忆任务 | 不接受 | 超出问题范围及当前内容 |
| 姓名栏与匿名计划冲突 | 接受 | 移除该栏并核对信息脚本 |
After removing a name field, what else should be checked? · 移除姓名字段后，还应检查什么？
Linked sections must remain consistent with the anonymous collection plan. · 相互关联的部分必须与匿名收集计划保持一致。
Close the review with a new check
- Give the reviewer the revised version and response log. Ask them to check whether the question, outcome and procedure now match. This is a focused follow-up, not a request to praise the entire project.
- Review a partner's one-page proposal using two substantive comments and one clear strength. Then revise your own draft and record one decision you did not accept, with a defensible reason. Keep all versions labelled so the final presentation uses the approved candidate.
Useful peer review is specific, criteria-based and respectful. A response log connects feedback with a justified revision and a check that the revision solved the problem.
以新的核查结束评审
- 向评审者提供修订版和回应日志。请他们检查提问、结果和流程现在是否一致。这是有针对性的后续跟进，而非请求对整个项目进行表扬。
- 使用两条实质性评论和一条明确优点评审同伴的一页提案。然后修订自己的草稿，记录一个未接受的决定及其有说服力的理由。保留所有带标签的版本，以便最终展示时使用经批准的候选版本。
有用的同伴评审应是具体的、基于标准的且尊重的。回应日志将反馈与有理由的修订及验证问题已解决的核查联系起来。