The AMBOSS analytics system can use your personal performance data to generate an adaptive question session. This helps you save time by focusing on the most relevant questions, targeting your individual knowledge gaps, prioritizing high-impact exam topics, and presenting questions in an order optimized for efficient learning. Here’s how a breakdown of how questions are selected.
Step 1: Apply relevant content filters
The system looks for questions matching your study objective.
Step 2: Prefer unanswered questions
The system narrows the matching questions down to those you haven’t answered yet. If it can’t find enough unanswered questions, it will recycle questions you have already answered. Question recycling is inspired by spaced repetition, with older incorrect questions resurfacing first.
Step 3: Build the session from the candidate questions
For each candidate question, the system estimates your chances of answering correctly given your current knowledge profile for the articles and competencies (e.g., pharmacotherapy, diagnosis) that question covers. The system then selects the question with the lowest estimated probability of a correct answer—in other words, the question where you appear weakest compared with the others.
After selecting a question, the system simulates its internal knowledge model as if you had answered that question correctly. This updated simulation influences the next selection step. As a result, the algorithm can move toward different weak areas instead of repeatedly targeting the same one.
This estimate–select–update loop continues until the session is fully populated. The process is similar to what is used in Computerized Adaptive Testing.
How to use it
The resulting adaptive session should target your weaknesses, while maintaining a mix of different topics and a reasonable difficulty level. If you are particularly weak in a certain area, you may get questions on it repeatedly–this is by design. By keeping your session sizes small (3-10 questions), you should notice that the algorithm guides you in a purposeful way through the curriculum, and over time you should see improvements in performance metrics, such as EPC.