A useful answer gives learning somewhere to go
A learner’s answer can open a conversation: what made that choice reasonable, which idea needs another look, and what should come next? QuestLab builds a shared practice experience around those questions.
Hosts bring a topic or their own lesson materials. Learners respond individually, examine explanations and discuss their reasoning. A shared follow-up round can then revisit concepts the group found difficult. The practical value is a connected workflow for practice, feedback and the next learning activity.
Our educational ambition is to help learners explain ideas more clearly, reconsider misconceptions and apply what they have learned in a new situation. The research below explains the design choices behind that ambition and how we plan to examine them.
Answer. Review. Discuss. Revisit.
Start with the material that matters
Generate a quest from a topic or use classroom notes, supported documents and images up to 10 MB. Text extraction and OCR turn supported printed text into material the host can review before generation. Rooms support 2–20 participants, including the host; learners join with a nickname and room code or QR route, without an account.
Make the feedback useful
Learners see accurate ideas they selected or missed, mistaken ideas they selected or avoided, and explanations. The host can ask what changed their thinking or invite them to defend a choice. Hosts choose standard or extra time when creating a quest, and speed adds no points.
Give the next round a reason
“Practice the gaps” creates shared follow-up around concepts the group needs to revisit. It gives the host a way to connect what happened in one round with what the group does next.
What this could look like in a lesson
Imagine a teacher bringing notes on photosynthesis. If learners share selections that suggest confusion between sunlight as an energy source and carbon dioxide as a source of matter, the feedback gives the group something specific to discuss. A follow-up round can revisit the distinction; a separate written explanation can check each learner’s application to a new example. This is an illustration of use, not a reported classroom study.
Research explains the design choices
Make a choice before seeing the answer
Retrieval research provides a reason to invite an active response before feedback. Roediger and Karpicke found benefits for delayed retention when learners recalled studied material. Their experiments included free recall; QuestLab uses statement selection. Our design hypothesis is that committing to a response activates knowledge and gives learners something to compare with the explanation that follows. [9]
Make feedback part of the activity
Butler and Roediger found that feedback after multiple-choice questions improved delayed correct recall and reduced the later use of incorrect alternatives. QuestLab pairs selections with explanations so learners can review the idea they missed and reconsider an attractive but mistaken answer. [10]
Make room for explanation
In undergraduate genetics courses, Smith and colleagues found that peer discussion combined with instructor explanation supported stronger performance on subsequent concept questions than either approach alone. This informs a facilitated use of QuestLab: learners explain their choices, compare interpretations and work through uncertainty with the host. [11]
Let evidence shape the next activity
Formative-assessment theory connects evidence with decisions about learning. QuestLab’s shared gap-focused practice gives that principle a place in the workflow: review the responses, choose what to revisit and try again. [5]
These connections form a research-informed rationale. Accurate questions, clear learning goals, suitable pacing and active facilitation are the conditions that make the cycle worth testing.
Make space for partial understanding
Learning progressions describe increasingly sophisticated ways of thinking within a domain. They direct attention to the ideas learners already use and the conceptual changes instruction can support. [1] [4]
That orientation matters in QuestLab. A learner can recognise one important relationship while still confusing another. Multiple statements and partial-credit feedback give the host material for exploring that pattern, then planning another opportunity to practise.
From a response to the next question
The assessment triangle connects a model of learning, observations from tasks and an interpretation of the evidence. Evidence-centered design asks which observations would justify a particular claim. [2] [3]
Recognising that one-half and two-quarters represent the same quantity is useful evidence. Explaining the relationship with a diagram and applying it to a new problem add different evidence. A quest can begin that sequence; a host can extend it with an explanation or application task.
Read the score alongside the response
| What you see | How to use it |
|---|---|
| 0–5 practice score | A summary of accurate and inaccurate selections on the question played. Use it to discuss that response and choose what to revisit. |
| −1 · Not Applicable | The host excluded a flawed or unsuitable question. It is omitted from progress calculations. |
| No evidence yet | A skipped or timed-out question, rather than a score of zero understanding. |
| Session averages | Descriptive summaries to read with the questions. Generated decks differ in content and difficulty. |
The 0–5 categories are response-performance indicators, not validated stages in a domain-specific learning progression. Partial-credit research supports investigating how scoring preserves information from multiple-response items. The particular rule and its intended interpretation still need their own evidence. [7]
AI helps author questions and proposed keys; a consistent rule summarises selections against the key. Hosts review source text and should examine questions and explanations, because generated content can contain plausible errors. They can discuss or exclude an unsuitable item. [8]
Use QuestLab for low-stakes formative practice, alongside explanations and other classroom evidence. Grades, placement and certification require assessments validated for those purposes. [6]
The pathway we expect to support
This logic model connects the experience with the learning changes we aim to achieve. Its arrows represent a proposed pathway to evaluate.
- 01 · Resources & conditions
Begin with a sound learning goal
A defined concept, reviewed questions and explanations, accessible devices, sufficient time and a facilitating host.
- 02 · Learning activities
Answer, review, discuss, revisit
Individual responses, explanatory feedback, facilitated discussion and shared gap-focused practice.
- 03 · Expected mechanisms
Activate, compare and explain
Use existing knowledge, reconsider an idea, articulate reasoning and focus the next attempt. [5] [9] [10] [11]
- 04 · Implementation outputs
Establish what took place
Attempts, feedback review, discussion and follow-up completion, recorded through session data and study instruments.
- 05 · Near-term outcomes
Clearer reasoning and useful next steps
More accurate explanations, fewer repeated misconceptions on new tasks and better-aligned teaching decisions.
- 06 · Later outcomes
Remember and apply
Retention at day 14 and application to unfamiliar examples, with broader conceptual development as a longer-term aim.
Delivery records establish whether the cycle took place; independent tasks establish whether learning changed. Discussion and explanation measures require study instruments alongside existing records. Findings should inform revisions to tasks, feedback and facilitation.
Building evidence that educators can use
As of 28 September 2026, no formal product validation or outcome study has been completed. The studies below are proposed; no active trial or preregistration is reported. Published research supports the design rationale. Product-specific learning effects and score validity are separate questions for evaluation.
The first study needs a defined concept, setting, learner group and responsible research lead. Educators should help confirm a realistic comparison, tasks should be reviewed and piloted independently, and a protocol should be preregistered before recruitment, with appropriate approvals and consent.
Five questions to guide the research
Does the experience support durable learning?
Compare facilitated QuestLab with structured teacher-led worksheet review: individual practice followed by normal solution review and explanation. Match the learning goal and time, account for baseline knowledge, and confirm the comparator with participating teachers. The prediction is stronger delayed performance with the full cycle. The primary outcome is an independent day-14 assessment with unfamiliar applications and blind-scored short explanations.
Do explanations help correct mistaken ideas?
Compare explanations with correct-answer-only feedback on matched questions and time. Measure independently coded misconception errors on unseen examples immediately and at one fixed, preregistered delayed endpoint. The prediction is fewer repeated errors with explanations.
What does discussion contribute?
Compare peer discussion with individual reflection after identical feedback, matching time and prompt demands. Measure individual explanation change before and after the activity and application to unfamiliar examples. Both groups receive the same final reviewed explanation afterward. The prediction is stronger reasoning and application after discussion.
Does follow-up help where it is needed?
Compare group-gap practice with broad review, matching time, question count and approximate difficulty. The prediction is greater independent improvement on concepts identified as difficult at baseline.
Can response patterns inform a better next step?
In a study configuration, randomly assign hosts to a proposed response-pattern report based on QuestLab feedback or a conventional score-and-answer summary for the same expert-reviewed cases, with equal planning time. Blind ratings assess how well proposed teaching plans address independently established learner needs. The prediction is better alignment; this study concerns planned decisions.
H1 is the first priority; H2–H5 are later component studies. Independent tasks, realistic comparisons and reporting of null or adverse results keep the model open to revision. Improvement only on repeated questions would challenge a claim of broader learning.
Study the pathway as well as the endpoint
H1 should collect individual explanations before feedback and after the full cycle, with matched measurement prompts in both groups. Blind coding can track changes in accuracy, reasoning and misconceptions. Relationships with day-14 performance can illuminate the pathway without establishing causal mediation.
Scoring rubrics should specify concept-level criteria, anchored examples and rules for ambiguous responses. Independent raters should train on separate practice answers and be blind to condition and measurement occasion where feasible. Report agreement and uncertainty. For H5, assess evidence interpretation, instructional alignment, subject accuracy and a follow-up check, allowing multiple defensible plans. [2] [3] [6]
The full whitepaper specifies the fixed retention point, study launch requirements, comparison conditions and analysis considerations, including classroom clustering and missing responses.
For education leaders considering implementation
ESSA’s “demonstrates a rationale” category connects a research-supported rationale with ongoing efforts to examine effects. Federal guidance describes a well-specified logic model with a study planned as part of implementation or underway elsewhere. [12] [13]
This paper provides a foundation for that conversation, not an ESSA rating or endorsement. Funding decisions require the applicable evidence and documented study arrangements. Section 1003 school-improvement funding requires one of the higher three evidence categories. [12]
For a first use, choose a concept learners have already encountered. Run a facilitated quest, ask each learner to explain one choice and use the responses to select the next activity. For implementation or evaluation enquiries, contact hello@questlab.app.
Keep the conversation open
QuestLab brings together teaching, STEM education research, biomedical research and learner mentoring experience. Our public research documentation explains the educational design, score meaning and evidence. Future study reports should include methods, findings and limitations; internal authoring instructions, production code and content-selection methods remain proprietary.
Read the full whitepaperReferences
- National Research Council (2007). Taking Science to School, chapter 8.
- National Research Council (2001). Knowing What Students Know, chapter 2.
- Mislevy, Almond & Lukas (2003). A Brief Introduction to Evidence-Centered Design. ETS RR-03-16.
- Alonzo & Steedle (2009). Developing and assessing a force and motion learning progression. Science Education, 93, 389–421.
- Black & Wiliam (2009). Developing the theory of formative assessment. Educational Assessment, Evaluation and Accountability, 21, 5–31.
- AERA, APA & NCME (2014). Standards for Educational and Psychological Testing.
- Betts, Muntean, Kim & Kao (2022; online 2021). Evaluating Different Scoring Methods for Multiple Response Items Providing Partial Credit. Educational and Psychological Measurement, 82, 151–176.
- NIST (2024). AI Risk Management Framework: Generative AI Profile. AI 600-1.
- Roediger & Karpicke (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17, 249–255.
- Butler & Roediger (2008). Feedback enhances the positive effects and reduces the negative effects of multiple-choice testing. Memory & Cognition, 36, 604–616.
- Smith, Wood, Krauter & Knight (2011). Combining peer discussion with instructor explanation increases student learning from in-class concept questions. CBE—Life Sciences Education, 10, 55–63.
- 20 U.S.C. § 7801(21). Definition of evidence-based, including the rationale category and school-improvement funding requirements.
- U.S. Department of Education (2016). Using Evidence to Strengthen Education Investments. See pages 8 and 12.
Version 1.4 · Updated 28 September 2026. A focused research synthesis, product design rationale and proposed evaluation roadmap.