Scoring & Assessment Rules
"Scoring" means different things by quiz type: Quiz focuses on right/wrong and totals, while Scored Quiz focuses on multi-dimensional analysis and reports.
Quiz: scoring and judgment
In a Quiz you can:
- Set points: Assign points to questions or correct options.
- Judge right/wrong: Mark correct answers; the system judges each question automatically.
- Sum totals: Points are added up automatically.
- Passing score: Set a pass threshold and show a pass/fail result.
- Answer feedback: Show respondents the correctness and explanation per question.
Points and right/wrong judgement are available in Quiz and Random Knowledge Quiz. Scored Quiz uses dimension formulas instead, and Outcome doesn't score at all.
Outcome: winning by votes
An Outcome quiz doesn't score — it counts votes. Each option votes for one or more outcome types, and the type with the most votes wins; ties go to whichever comes first in the outcome list.
Every outcome can carry its own CTA button, wiring an entertaining result straight into a next step. See The four quiz types.
Scored Quiz: dimensional analysis and AI evaluation
In a Scored Quiz, scoring is more about "profiling" than right/wrong:
- Multi-dimensional scoring: Map questions to several dimensions; after submission, each dimension gets a score, often shown as a radar chart.
- Result profile: Produce personality/ability conclusions based on dimension scores.
- AI evaluation: AI turns the total score, level, and dimension scores into suggestions and commentary (it does not read the raw answer text).
Which questions are scored
A Scored Quiz total comes only from the formula, which references question codes. Only questions that yield a number can be referenced: choice questions (via per-option scores), rating, number, date, time, and cascade.
Fill-in-the-blank only collects text and is never scored — free text cannot take part in arithmetic, so it produces no score and does not appear among the insertable fields in the formula editor. That makes it the right way to ask one open-ended question ("anything else we should know?") inside a scored assessment.
With no formula the total score is empty, no level can ever match, and the overall-assessment block disappears from the result page. Whenever you configure levels, configure the formula too.
About AI evaluation quota
AI evaluation consumes your team's AI quota. Quota is counted per cycle; once exceeded, AI evaluation is skipped for that cycle (basic scoring is still generated). See AI Credits & Billing for billing, monthly resets, and top-ups.
Levels and result-page actions
Split the total score into levels (0–60 needs work / 60–80 good / 80+ excellent). Each carries a name, colour and description, and can hold a CTA button — the low band might offer "Book a consultation". This is the step that wires an assessment into conversion.
Don't want respondents seeing a number? You can hide the overall score and keep only the level name, description and summary — better for quizzes that should feel like a diagnosis rather than an exam.
Report generation
- Synchronous: Basic scoring is calculated immediately after submission.
- Asynchronous: Heavier computation such as dimensional analysis and AI evaluation is generated asynchronously; reports go through a "generating → completed" state, and the system retries automatically on failure.
For viewing, exporting, and delivering reports, see Reports & AI Evaluation.
Conditional Logic
Conditional logic shows or hides later questions based on earlier answers, turning a fixed questionnaire into an adaptive one. How a rule reads, and why it is Scored Quiz and Outcome only.
Form Settings
Everything about a quiz that isn't a question: title, description and cover image, the open and close switches, respondent sign-in, result visibility, and the fields that feed search indexing.