引導式測驗 · 中等 等級
AI 線上討論板測驗
使用人工智慧設計和促進線上討論,同時保留學生推理、包容性參與、隱私和透明的人工智慧使用規則。
問題 1 的 8
Which task can AI assist with when planning an online discussion?
預覽每個問題8 問題
當您參加測驗時,就會出現答案和解釋。
- Which task can AI assist with when planning an online discussion?
- Guarantee no student can use AI
- Replace course learning goals
- Generate prompt drafts for instructor revision
- Prove academic misconduct from writing style
- Which prompt design can make student reasoning more visible?
- Ask for a generic summary of a broad topic
- Ask students to compare evidence from a specific course source
- Require a long post without a rubric
- Ask the model to write the response
- Why should an instructor review an AI summary of a discussion thread?
- It always includes every post exactly
- It creates a validated grade automatically
- It removes all personal information
- It may omit minority views or overstate agreement
- How should AI-detection output be treated in an academic-integrity concern?
- A detector score alone is not proof of misconduct
- All detectors identify every AI-generated text
- A high score establishes intent
- Detectors replace institutional procedures
- How should a course explain student AI assistance?
- Leave expectations implicit
- Promise that AI use cannot be noticed
- State which uses are allowed and how to attribute them
- Prohibit discussion of tools in every context
- What privacy step should an instructor take before summarizing student posts with an external tool?
- Upload the entire class archive by default
- Use an approved service and minimize shared student information
- Assume discussion posts are public
- Rely on students to remove identifying details after the full archive is uploaded
- Which rubric is more useful for evaluating discussion quality?
- One that grades only post length
- One that uses AI detector scores
- One that rewards posting the most times
- One that distinguishes evidence, reasoning and engagement
- What should guide the use of automated moderation or grading?
- Disclosure, a reviewable rubric, instructor oversight and correction paths
- Silent automatic decisions
- Only the number of flagged posts
- A model's confidence without context