引導式測驗 · 硬 等級
醫學生人工智慧和 USMLE 準備測驗
測試如何審查人工智慧編寫的 USMLE 問題、NBME 專案寫作規則(例如覆蓋選項)以及如何將人工智慧與現有的學習資源結合使用。
問題 1 的 8
自 2022 年 1 月以來,USMLE 第 1 步的報告如何?
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。
- 自 2022 年 1 月以來,USMLE 第 1 步的報告如何?
- As a three-digit score
- As pass/fail
- As an optional exam
- It was merged into Step 2 CK
- 為什麼人工智慧在 USMLE 式議題上的強勁表現卻無法使其自己的練習題達到考試品質?
- The studies used only pediatric questions
- AI performance studies were retracted
- AI cannot write in a vignette format
- Answering questions well does not guarantee that the questions or explanations it writes are correct
- NBME 覆蓋選項規則對寫得好的問題有何要求?
- A prepared student can answer from the vignette and lead-in without seeing the choices
- The answer options are hidden until the timer ends
- The question has exactly two options
- The options cover every organ system
- 學生應該如何處理人工智慧編寫的有兩個合理答案的小插圖?
- Memorize the keyed answer anyway
- Choose the longer option
- Discard or fix the item instead of learning from it
- Report it to the NBME
- 該指南推薦了哪些資源來判斷步驟 1 的準備?
- The number of AI questions completed
- NBME self-assessments
- A chatbot's estimate of your score
- Social media study groups
- Which technique does the guide describe for checking whether an AI question is keyed correctly?
- Ask the same chat to confirm its answer
- Count the words in each option
- Look for the option with the most technical term
- Answer the question blind in a fresh chat and investigate any disagreement with a trusted resource
- What sycophancy risk does the guide warn about when studying with AI?
- The model may drop a correct answer when the student pushes back
- The model refuses to answer medical questions
- The model always asks for payment
- The model deletes the student's notes
- Which error mode in AI-made USMLE questions relates to clinical guidelines?
- Vignettes that are too short
- Options listed alphabetically
- Relying on outdated or non-US guidelines
- Using lab values with units