Ikibazo gihuza ibibazo · Hagati Urwego

AI kubitsinda ryabanyeshuri hamwe nimbonerahamwe yo Kwicara

Koresha AI kugirango ubyare amatsinda yo mwishuri mugihe wubaha imbogamizi zigaragara, icyubahiro cyabanyeshuri, kugerwaho, gusubiramo abarimu no kwiherera.

Ikibazo 1 Bya 8

What should a teacher specify before asking AI to suggest student groups?

Reba buri kibazo8 ibibazo

Ibisubizo nibisobanuro bigaragara mugihe ufata ikibazo.

  1. What should a teacher specify before asking AI to suggest student groups?
    • Rank students only by prior test scores
    • Use fixed labels such as high and low ability
    • The lesson objective and relevant constraints
    • Include diagnoses by default to simplify sorting
  2. Why should AI not infer sensitive traits from grades or free-text notes?
    • Grades perfectly identify disability
    • Such inferences can be inaccurate, stigmatizing, or unauthorized
    • The model can directly observe classroom relationships
    • Sensitive traits are required for every group plan
  3. Which classroom requirement is a hard constraint rather than a soft preference?
    • A preferred seat neighbor
    • A teacher preference for a discussion mix
    • An optional role assignment
    • A requirement such as honoring an accessibility need
  4. What should happen if the requested constraints cannot all be satisfied?
    • The tool should surface the conflict for teacher review
    • It should silently drop an accommodation
    • It should exclude a student without explanation
    • It should infer which student matters least
  5. Which check helps prevent a malformed seating plan?
    • Verify only total headcount, not whether each student is assigned once
    • Assume unlisted students are absent from the class
    • Verify each student appears once and group sizes are valid
    • Check only that groups have the same average test score
  6. Why generate multiple candidate groupings?
    • The first option always violates policy
    • Teachers can compare tradeoffs against classroom context
    • Students should never be grouped
    • More candidates guarantee perfect fairness
  7. What check should precede uploading real student data to a third-party tool?
    • Upload full records to test the interface
    • Assume student names are public
    • Disable all review
    • Check school policy, authorization, and data minimization
  8. Who should make the final classroom grouping decision?
    • The educator using the plan and relevant context
    • The model without review
    • The software vendor alone
    • An automated ranking of students