Recorded practice
A server-checked quiz attempt records answers, score, and completion time. The answer key is not trusted from the browser.
Evidence Center · revision 24 September 2026
Clear about what we teach,
what we know, and what we do not.
Skill X is a learning product, not a claim of guaranteed outcomes. This page explains the learning loop, the content checks, the correction path, and the limits of our current evidence so families and educators can make an informed choice.
01 · Learning design
How Skill X teaches
Lessons move from a visible idea to an explanation, a worked example, practice, and later retrieval. The sequence is designed to make reasoning observable, not to replace a school teacher or official syllabus.
1
See
A visual model or concrete example gives the learner something to notice before symbols become dense.
2
Explain
“Why” prompts and vocabulary connect the observation to a concept. Learners can write their own explanation privately.
3
Solve
Worked examples reveal one decision at a time, then ask the learner to attempt a related problem.
4
Retrieve
Memory Check asks for an answer before showing the lesson again. Confidence is recorded as a reflection, not a grade.
5
Transfer
An optional “Transfer it” prompt invites a home, school, nature, technology, or community connection.
Research boundary: the design is informed by established learning-science literature, but the sequence itself is not a product-specific efficacy study.
02 · Honest progress
What mastery means here
Opening a page, earning XP, or selecting “I get it” is not mastery evidence. Skill X uses a narrower, descriptive contract.
Delayed recall
A topic is labelled “remembered over time” only after two successful delayed Memory Checks. “Time to remember” means the evidence is incomplete.
Parent view
Reports distinguish server-derived quiz activity from learner reflections and parent-authored notes. They are indicators, not predictions.
03 · Review method
How review works
Secure review sessions interleave topics from a learner’s completed work. The browser receives a session question set, while answer evaluation and progression remain server-authoritative when an account is available.
Memory Check first: the lesson stays hidden until the learner commits an answer. A confidence reflection follows, so a correct guess is not silently treated as secure understanding.
04 · Content review standard
What we check before a release
Every learner-visible lesson carries a content release marker and a structured-check status in the lesson view. The pipeline checks shape and safety fields; it does not replace qualified human curriculum sign-off.
Structure
Title, summary, outcomes, worked examples, flashcards, and quiz references must point to the same topic and class band.
Assessment quality
Automated gates look for answer-position patterns, repeated stems, template distractors, and metadata mismatches.
Safety and age fit
Activities include age-aware wording and an adult-supervision reminder where hands-on materials or household items are involved.
Still required: a qualified educator must sign off factual accuracy, pedagogy, cultural fit, and age appropriateness before a high-stakes public curriculum claim is made.
05 · Safety by age
Safety is part of the lesson contract
Class bands guide the language and examples. They do not remove the need for an adult to decide whether an activity is suitable for a particular child.
1–4
Concrete foundations
Short instructions, familiar objects, observation-first tasks, and no assumption that a child can handle tools or heat independently.
5–7
Guided abstraction
Models and variables become more explicit. Adult guidance remains appropriate for experiments and household materials.
8–10
Formal models
Equations and evidence interpretation increase, while activities still require the safety directions on the lesson and local supervision.
06 · Corrections policy
Report a problem; we keep an audit trail
Inside a lesson or preview, choose Report an issue. The report is saved on the current learner profile and can open a pre-filled email draft to contact@smaraze.com. It is not described as received by the review team until delivery succeeds.
1 · ReportedSpecific topic, category, urgency, and content version are captured.
2 · TriagedFactual, pedagogy, access, and safety reports are prioritised.
3 · ReviewedAn educator or content owner checks the source and learner impact.
4 · CorrectedThe source lesson is updated, tested, and assigned a new release marker.
5 · PublishedOnly after QA is the corrected lesson included in the next public bundle.
High-severity handling: a safety or misleading-content report is an operator decision point for holding a lesson from future publication; the browser does not pretend that local storage is a central quarantine system.
07 · Learning-science references
Where the design ideas come from
These sources inform design choices; they are not endorsements of Skill X and do not establish product-specific outcomes.
- Dunlosky et al. (2013). Improving students’ learning with effective learning techniques. Psychological Science in the Public Interest.
- Roediger & Karpicke (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Journal of Memory and Language.
- Bjork (1994). Memory and metamemory considerations in the training of human beings. In Metacognition.
08 · Pilot results
No product-specific results claimed yet
Skill X does not currently publish a learner-outcome study, retention benchmark, testimonial count, or “100%” claim. When a pilot is run, we will document its cohort, consent process, measures, dates, limitations, and analysis before using the results in public messaging.
09 · Language and context
Learning can start from a familiar place
Lesson screens now expose a locale-readiness control for English (India), Hindi, Odia, Bengali, Tamil, and Telugu. English is the currently reviewed content language; other locales are clearly marked as translation in progress until human language and curriculum review is complete.
Stable learning record
Changing the interface language never changes a concept ID, progress record, mastery evidence, entitlement, or report metric.
Optional India-first examples
Lessons may offer a market, monsoon, local plant, cricket, map, space, or public-technology connection. Families can choose another example; no single city, income, or household is assumed.
Safety stays explicit
Hands-on connections remain observation-first and point to adult supervision where heat, pressure, chemicals, tools, electricity, traffic, or weather can create risk.
10 · Transfer and healthy breaks
Learning can travel beyond the screen
After a secure quiz, an optional real-world challenge asks the learner to separate useful information, choose a strategy, answer, and explain their reasoning. A parent-controlled screen-time break may offer a safe off-screen prompt.
Transfer evidence
Challenge attempts are descriptive evidence of strategy, answer, and explanation. They never unlock lessons, change entitlement, or silently become mastery.
Explored, not mastered
Finding objects, observing a shadow, measuring safely, or comparing a container can be recorded as explored-only activity. It is not a score or a streak obligation.
Parent-aware safety
Some prompts include an optional parent note and an adult-supervision reminder. Families decide whether an activity is suitable and stop if anything feels unsafe.
11 · Known limitations
What this release does not promise
Human curriculum sign-off, real-device accessibility review, production load testing, and legal review are separate release gates.
Transactional email delivery is not connected in the local contract; recovery messaging must not claim that an email was sent until the binding and delivery test exist.
Live Razorpay payment and webhook verification remains an external test gate. Local pricing copy is not proof of a successful payment.
Local correction reports and learner artifacts are private browser state unless the documented account/report service is used. They are not a substitute for backup.