Quiz połączony z przewodnikiem · Średni Poziom

Quiz dotyczący sztucznej inteligencji dla pacjentów i materiałów edukacyjnych

Testuje docelowe poziomy czytania, granice formuł czytelności i zasady przeglądu ręcznego w przypadku tłumaczonych maszynowo materiałów informacyjnych dla pacjentów.

Powiązane ścieżki prowadząceAI dla materiałów edukacyjnych dla pacjentów
Pytanie 1 z 8

What reading level do health literacy guidelines commonly recommend for AI-drafted patient handouts?

Podgląd każdego pytania8 pytania

Odpowiedzi i wyjaśnienia pojawiają się w miarę rozwiązywania quizu.

  1. What reading level do health literacy guidelines commonly recommend for AI-drafted patient handouts?
    • College level
    • Tenth to twelfth grade
    • Sixth to eighth grade
    • Second grade
  2. What is a key limitation of readability formulas like Flesch-Kincaid when checking an AI handout?
    • They measure sentence and word length, not accuracy or cultural fit
    • They only work on printed documents
    • They need a physician's license to run
    • They measure accuracy but not length
  3. What does AHRQ's PEMAT evaluate in patient education materials?
    • Printing cost and paper quality
    • Understandability and actionability
    • Search engine ranking
    • Copyright status
  4. Why should the prompt tell the AI to use only a vetted source?
    • To make the handout longer
    • To raise the reading grade level
    • Because AI cannot read guidelines
    • To keep the model from adding unverified or outdated content
  5. Under the 2024 Section 1557 rule, what is required when machine translation is used for critical content?
    • Nothing, because machine translation is exempt
    • Only a disclaimer at the bottom of the page
    • Review by a qualified human translator
    • Approval by the software vendor
  6. Why does the guide suggest expressing risk as '1 in 10 people' instead of a percentage?
    • Many readers find natural frequencies easier to understand
    • Percentages are banned in patient materials
    • Natural frequencies lower the Flesch-Kincaid score to zero
    • AI models cannot calculate percentages
  7. Why should AI translations be back-translated and tested with native speakers?
    • Because AI translation is always perfect
    • Translation quality varies and is weaker for languages with less training data
    • Because back-translation shortens the text
    • To qualify for a readability certificate
  8. Which is an example of simplification turning into inaccuracy?
    • Replacing 'edema' with 'swelling in your legs or feet'
    • Using headings and short bullets
    • Writing sentences addressed to 'you'
    • Dropping 'unless your doctor told you otherwise' from an instruction