Workflow model
Quiz branching and scoring decision tree: from answers to outcomes
A reusable decision tree and truth-table checklist for choosing scores, branches, outcomes and follow-up rules.
- Published
- 24 August 2026
- Reading time
- 12 min
- Evidence
- Decision tree and logic truth table
- Author
- The AI Quiz Lab
Use scoring when answers accumulate toward a measurable dimension. Use branching when one answer changes the next relevant question. Use outcomes when the final result must explain what the pattern means. Add follow-up only after those rules are complete and testable.
Key findings, 24 August 2026
- Scores, branches and outcomes solve different logic problems.
- A truth table is the fastest way to expose gaps, overlaps and unreachable outcomes.
- Every possible response path needs a defined end state, including ties and missing answers.
What is quiz branching?
Quiz branching is conditional routing that changes the next question or section based on a previous response. Scoring assigns values across answers. Outcomes translate the resulting pattern into a result. A quiz can use one, two or all three, but they should not be treated as interchangeable.
Which logic model should you choose?
| Decision | Use | Avoid when |
|---|---|---|
| Direct answer | One response determines the result | Several signals must be combined |
| Weighted score | Answers add to one or more dimensions | The next question depends on a gate |
| Branch | An answer changes what is relevant next | All participants need the same evidence |
| Rule-based outcome | Several conditions define a segment | Rules overlap or lack a fallback |
| Follow-up workflow | A verified outcome triggers a next step | Consent, ownership or error handling is unclear |
The reusable decision tree
- Can one answer determine the result? If yes, use a direct branch and document the exception path.
- Do answers measure a shared dimension? If yes, assign a transparent score and define the range.
- Are several dimensions independent? If yes, score each separately before choosing an outcome.
- Should an answer skip irrelevant questions? If yes, add a branch, then verify that skipped items do not affect scoring.
- Can rules tie or overlap? If yes, set a tie-breaker or neutral fallback.
- Does the result trigger an action? If yes, define consent, owner, timing, failure handling and a non-automated fallback.
Worked model: a training needs quiz
Assume four questions measure confidence and recent experience. Confidence answers score 0 to 2. A “no recent experience” answer opens one extra context question. Outcomes are Practice basics for 0 to 2, Guided application for 3 to 5, and Independent challenge for 6 to 8.
| Case | Score | Branch | Expected outcome |
|---|---|---|---|
| All lowest | 0 | Extra context shown | Practice basics |
| Boundary low | 2 | As answered | Practice basics |
| Boundary middle | 3 | As answered | Guided application |
| Boundary high | 6 | As answered | Independent challenge |
| Maximum | 8 | No irrelevant detour | Independent challenge |
| Missing answer | Incomplete | Return to required item | No outcome yet |
What should the logic specification contain?
Reusable logic specification
List every question ID, response option, score contribution, branch condition and required status. Define outcome ranges with inclusive boundaries. Add tie, missing-answer and no-match behavior. Produce a truth table covering minimum, maximum, every boundary, every branch, every outcome and every fallback. Do not create follow-up actions until all paths resolve exactly once.
What failures should the truth table catch?
- Gap: a score or answer combination has no outcome.
- Overlap: one path satisfies two outcomes.
- Unreachable result: no possible answers can produce the configured range.
- Score contamination: a skipped question still contributes a default value.
- Branch loop: routing can send a participant back indefinitely.
- Premature automation: a partial or corrected response triggers follow-up twice.
What are the limits?
A clean logic model does not prove that the quiz measures a valid construct or that an outcome is fair. Avoid using self-built quiz logic for high-stakes decisions. For ordinary education, training or content personalization, disclose what the result means and let a human review consequential follow-up.
Use this alongside the question quality checklist, the six-part prompt specification and the AI Quiz Lab methodology.
When questions are source-generated, pair the logic test with the source-fidelity audit. When answer options are generated, use the distractor QA test.
Sources and verification
This guide was verified on 24 August 2026. Facts are linked to primary or institutional sources. The worked models and checklists are original AI Quiz Lab evidence assets, designed to be repeated with any generator.
If a source changes or you find an error, use our correction path. The publication-neutral next step is to test the framework on a small, low-stakes quiz before using it at scale.