# AI Assessment for Open-Ended Questions

Automatically evaluate learner responses to open-ended questions and provide
instant feedback based on accuracy.


SETUP PROCES

 1. Add an open-ended question

 2. Enable AI Assessment: Check "This open-ended question has a correct answer"
    
    
    
    

 3. Provide an "Example of a correct answer": Write an example of what a correct
    response should include

 4. Create Feedback Messages: Write two responses - one for correct answers, one
    for incorrect answers


AI EVALUATION PROCESS

 * AI compares learner responses to your example in real-time

 * Uses semantic matching (understands meaning, not just exact words)

 * Considers lesson content and training data for context

 * Classifies responses as "correct" or "incorrect"

 * Automatically sends appropriate pre-written feedback to learners

 * Results count toward overall accuracy calculations


BEST PRACTICES FOR QUESTIONS

✅ Good for:

 * Questions with objectively correct answers (e.g., "Name three countries in
   Africa")

 * Scenarios with multiple correct approaches but clear wrong answers

 * Knowledge checks with definitive criteria

❌ Not suitable for:

 * Creative or purely subjective responses

 * Opinion-based questions


BEST PRACTICES FOR WRITING EXAMPLES OF CORRECT ANSWERS

Use conversational, guiding language:

 * "Some examples might include..."

 * "The answer should focus on x, y, z"

 * "Key words the user might mention include..."

Provide criteria and multiple acceptable examples rather than one perfect
answer.


ADMIN CONTROLS

As an admin, you can go to Analytics > Select a Course > Engagement > Responses
to:

 * View individual learner responses and exact text submitted

 * See correct/incorrect percentages like multiple choice questions

 * Manually override AI-evaluated decisions



--------------------------------------------------------------------------------



--------------------------------------------------------------------------------