Skip to content

ENGINEERING_CURRICULUM.md

Latest commit

 

History

History
238 lines (182 loc) · 6.8 KB

ENGINEERING_CURRICULUM.md

File metadata and controls

238 lines (182 loc) · 6.8 KB
Raw

Engineering Curriculum Engine

This pass adds a kid-first Grade 1 through Grade 4 engineering and systems layer for the learning gate app.

The design goal is:

  • building over memorizing
  • noticing over jargon
  • cheap materials over kits
  • failure as information, not as a wrong-answer moment
  • design-build-test-improve cycles that feel concrete enough for a first grader

Primary structured files:

  • curriculum/engineering-curriculum.js
  • curriculum/diagnostics/engineering-placement.js
  • data/engineering/engineering-skill-graph.json
  • data/engineering/engineering-lessons.json
  • data/engineering/engineering-tasks.json
  • data/engineering/engineering-mastery.json
  • data/engineering/engineering-rewards.json
  • data/engineering/engineering-parent-reporting.json

What Is In The Data

The engineering layer now includes:

  • a prerequisite-linked skill graph from Grade 1 through Grade 4
  • a short engineering placement mini-diagnostic
  • a daily quick-prompt plus weekly build-quest loop
  • a mastery model that scores habits, not only correctness
  • companion reward hooks tied to prediction, testing, noticing, improving, and explaining
  • parent-facing reporting fields and sample summary outputs
  • 24 seeded lessons
  • 24 lesson task banks
  • 8 tasks per lesson
  • 192 seeded engineering tasks total

Skill Graph

The roadmap is organized as a skill graph, not chapter-only progression.

Core domains:

  • observe how something works
  • parts and wholes
  • cause and effect
  • structure and stability
  • strong vs weak shapes
  • balance and weight
  • push, pull, and motion
  • friction, ramps, wheels and axles
  • levers
  • pulleys as stretch
  • materials and properties
  • measurement for building
  • constraints and tradeoffs
  • prototype, test, failure-as-information, improve
  • compare designs
  • systems as input, process, output
  • energy at a kid level
  • simple circuits as stretch
  • robotics and automation connection as stretch

Grade emphasis:

  • Grade 1: concrete objects, support parts, ramp motion, bridges, towers, material matching
  • Grade 2: balance, friction, wheels, measurement, failure points, version two thinking
  • Grade 3: fair tests, constraints, tradeoffs, levers, systems, energy paths
  • Grade 4: prototype planning, multi-constraint design, iteration records, prototype comparison, circuits/automation stretch

Placement Mini-Diagnostic

The placement pass is intentionally short and companion-framed.

Target length:

  • expected: 8-10 tasks
  • hard max: 12 tasks

Included task types:

  • identify parts of a system
  • predict what happens if something changes
  • choose the stronger shape
  • choose the best material
  • spot why a tower fell
  • choose how to test fairly
  • pick the better improvement
  • compare two designs
  • choose a constraint-aware solution
  • identify input in a simple system

Placement output includes:

  • strongest engineering habits
  • shaky engineering habits
  • recommended first build quests
  • parent-facing summary
  • practice needs in observation, prediction, building, testing, measurement, or iteration

Daily And Weekly Loop

Engineering should usually appear in one of two forms:

  • 2-5 minute quick prompt
  • 10-25 minute weekly build/test quest

Loop:

  1. problem
  2. plan
  3. build or imagine build
  4. test
  5. notice what happened
  6. improve
  7. explain

Rules encoded in engineering-mastery.json:

  • ask for a prediction before tests and meaningful changes
  • ask for measurement when a compare claim needs evidence
  • ask what changed? after each version shift
  • reward careful testing even if the build does not work
  • reward version two more strongly than lucky version one success
  • stop early into observation/talk mode before frustration spirals
  • support offline imagine-build mode when materials are unavailable

Mastery Model

Engineering is not scored as only right or wrong.

Statuses:

  • introduced
  • practicing
  • builder
  • tester
  • improving
  • stretch-ready
  • needs_support

Scored dimensions:

  • observation
  • prediction
  • explanation
  • use of evidence
  • testing fairly
  • iteration / improvement
  • measurement
  • persistence
  • compare designs

Advancement uses:

  • repeated evidence across sessions
  • whether the child predicts before testing
  • whether a second version is purposeful
  • whether explanations use because/evidence language
  • whether fair-testing habits are present when required

Reward Hooks

The companion reward layer is deterministic and effort-based.

Reward moments:

  • smart prediction
  • careful observation
  • good test
  • build completed
  • failure noticed
  • improvement made
  • stronger second version
  • clear design explanation
  • weekly engineering quest complete

No random reward design is used.

Parent Summary Model

The parent-facing model supports:

  • today completed or not
  • engineering skill practiced
  • build or quest attempted
  • materials used
  • prediction
  • test result
  • improvement
  • explanation quality
  • strongest engineering habits
  • shaky engineering habits
  • suggested offline extension
  • safety note if needed
  • practice emphasis

UX Recommendations

Engineering should stay kid-friendly:

  • one challenge at a time
  • clear material list first
  • read-aloud always available
  • optional simple timer only for longer quests
  • future photo/upload placeholder for build evidence
  • what changed? reflection after testing
  • no wrong-answer framing for failed builds
  • parent mode for offline build help and safety support

Cross-Subject Fit

Engineering is a cross-subject layer, not a separate silo.

  • Math: measurement, comparison, shape reasoning, count-based test evidence
  • Coding: input/process/output, debugging, iteration, automation stretch
  • Creativity: inventing, prototyping, materials, version-two redesign
  • Thinking: prediction, causality, tradeoffs, evidence-first explanation
  • Real life: toys, furniture, packing, simple repairs, ramps, carts, lights, tools

Developer Wiring First

Wire these first:

  1. curriculum/diagnostics/engineering-placement.js for onboarding or subject unlock logic
  2. data/engineering/engineering-lessons.json plus engineering-tasks.json for daily queue generation
  3. data/engineering/engineering-mastery.json for habit-based scoring
  4. data/engineering/engineering-rewards.json for companion reactions
  5. data/engineering/engineering-parent-reporting.json for parent summaries

Open Questions For Bailey

  • Should engineering appear every day as a quick prompt, or only on selected days plus a weekly quest?
  • Should build evidence stay optional text-only for phase one, or do we want a real photo placeholder now?
  • Do we want engineering unlocked from day one, or only after onboarding reading/math placement is complete?
  • Should offline build quests count as full gate progress when the child explains the test well but does not upload evidence?
  • For stretch content, should circuits remain picture-only in phase one unless a parent opts in?