Question loop

Once you have a LessonSession from lessons.start(), the question loop is how your app drives the lesson forward. Call session.next() to get a question, render it, then call session.submit() with the student’s answer.

The basic loop

let question = await session.next()

while (question) {
    // Render the question, collect an answer from the student
    const answer = await renderQuestion(question)

    const result = await session.submit({
        question: question.id,
        response: answer,
    })

    // result.score, result.correct, result.complete are available

    question = await session.next()
}

// No more questions -- complete the lesson
const completion = await session.complete()

Fetch a question

session.next() returns the next LessonQuestion, or null when there are no more questions:

const question = await session.next()

if (!question) {
    // Lesson is complete, call session.complete()
}

Each normalized question includes an id; content and content.rawXml are optional in the contract. Check for XML before rendering and handle unsupported or missing content in your UI. Your app supplies a QTI renderer.

See the reference for the full type.

Select a renderer that supports the item interaction and response format. The managed-session submit() accepts a string and the server sends it as responses.RESPONSE; it does not expose every multi-response QTI contract. Use the returned grading result for feedback.

Submit an answer

After the student answers, submit their response:

const result = await session.submit({
    question: question.id,
    response: selectedAnswer,
})

The result tells you whether the answer was correct, the updated score, and whether the lesson is now complete. After each submission, score updates and a complete result stops subsequent next() calls. The public finalized property updates only after complete().

See the reference for full parameter and return type documentation.

const result = await session.submit({ question: question.id, response: 'A' })
console.log(result.correct) // Whether the answer was correct
console.log(session.score) // Updated cumulative score
console.log(result.complete) // Whether the answer ended the question loop

Adaptive vs linear delivery

The SDK handles two delivery modes transparently. Your code uses the same next() / submit() loop regardless of mode.

You do not need to detect which mode is active. The LessonSession handles it based on the lessonType set during lessons.start().

Adaptive (`powerpath-100`)

Each next() call hits the server, which asks PowerPath for the next question based on the student’s performance so far. Questions are served one at a time, and difficulty adapts.

Linear (`quiz`, `test-out`, etc.)

On the first next() call, the SDK fetches all questions at once and buffers them locally. Subsequent next() calls return from the local buffer without a network round trip. Already-answered questions are skipped automatically.

Track progress during the loop

The session properties update in real time as the student progresses. Use the returned result.complete and session.score to render progress. These are mutable object properties; copy returned values into your framework state when each request resolves. For linear lessons, submit returns correct: false and score: 0 before finalization; do not show that placeholder as a scored wrong answer.

Getting Started

CLI

SDK

Reference

The basic loop

Fetch a question

Submit an answer

Adaptive vs linear delivery

Adaptive (`powerpath-100`)

Linear (`quiz`, `test-out`, etc.)

Track progress during the loop

Next steps

Completion

Reference

Getting Started

CLI

SDK

Reference

​The basic loop

​Fetch a question

​Submit an answer

​Adaptive vs linear delivery

​Adaptive (powerpath-100)

​Linear (quiz, test-out, etc.)

​Track progress during the loop

​Next steps

Completion

Reference

The basic loop

Fetch a question

Submit an answer

Adaptive vs linear delivery

Adaptive (`powerpath-100`)

Linear (`quiz`, `test-out`, etc.)

Track progress during the loop

Next steps