怎么用 AI 出题与批改[object Object]
微型指南 · 约 7 分钟Guide · about 7 min · 引用核验 2026-11-19(仅引用链接与日期) · refs checked 2026-11-19 (reference links and dates only)
让 AI 出「能诊断理解」的题,而不是背答案的题;批改给反馈,不只给分数。Generate questions that diagnose understanding, not recall; grade with feedback, not just a score.
先给知识点、认知层级(记忆/理解/应用/分析)和题型比例,再让 AI 出题。
每题必须带答案、评分点、常见错误和对应知识点;没有评分点的题不要用。
让 AI 生成「干扰项解释」:每个错误选项对应哪种误解,方便讲评。
批改主观题时先给评分 rubric,再让模型按点给分并引用学生原句;人复核边界分。
把 AI 批改结果当作「初评」而不是终评,尤其是作文、开放题和涉及价值的题目。
定期抽查 AI 出题的正确性;模型会编出看似合理但条件矛盾的题。
Specify knowledge points, cognitive level (recall/understanding/application/analysis) and item mix before generating.
Every item needs an answer, scoring points, common errors and mapped knowledge point. Skip items without a rubric.
Ask for distractor explanations: which misconception each wrong option represents.
For open responses, give the rubric first; have the model score by point and quote the student’s sentence; a human checks borderline cases.
Treat AI grading as a first pass, especially essays, open questions and values-related items.
Audit generated items regularly; models invent plausible questions with contradictory conditions.
例子与判断Example and interpretation
以下是教学情境,不是本站运行或实测记录。A teaching scenario, not a measured or executed site test.
让模型出 5 道选择题,每题带答案、评分点、干扰项解释;你挑出 1 道条件矛盾的题,改成可判断的版本,再用 rubric 批 3 份学生答案。Ask for five multiple-choice items with answers, scoring points and distractor explanations; find one contradictory item, rewrite it, then grade three student answers against the rubric.
做完例子后怎么检查How to check your practice
先用自己的话解释上面的判断,再检查输入条件,最后回原文验证。说不清原因时回概念;缺日期、字段或权限时补输入;结论无出处时先撤回结论。Explain the example in your own words, inspect input constraints, then verify against the source. Revisit concepts if you cannot explain why; supply missing dates, fields or permissions; withdraw unsupported conclusions.
按这篇主题,先从“概念没分清”练习查起For this topic, start with concept confusion · 查看全部六个情境与理由See all six scenarios and their reasoning