Instructional Design · Platform Development · UX Audit

Designing an inquiry system that connects air quality data to K–12 civic reasoning.

AirBioCivic is a research-driven instructional design project combining a 6-module SSI curriculum, a live sensor data platform, and an instructor support system — all designed to turn real environmental data into civic inquiry for K–12 classrooms.

SSI PedagogyNeeds AnalysisUX AuditTech Design LeadFull-Stack Dev
ABC Platform
Heat Map
Analysis
Raw Data
Mean PM2.5
18.4
Good
Max
41.2
Moderate
Groups
4
Active
PM2.5 by GroupWHO ref: 15
G1G2G3G4
Role
Tech Design Team Lead
(2-person sub-team)
Context
Tamgu Lab · TC Columbia
Multi-team project
Deliverables
Platform · 6-module curriculum · Instructor Guide · UX Audit
Audience
K–12 Students &
Classroom Instructors
01 · Problem

The gap between science class and real-world reasoning.

K–12 environmental science instruction often treats data as something students receive rather than something they act on. Students learn about air quality in textbooks, but rarely encounter its local, civic dimensions — who is affected, why it matters in their neighborhood, and what they might do about it.

How might we design a learning experience that connects real environmental data to student agency — turning passive consumption into active, civic-minded inquiry?

The design challenge was twofold: build a curriculum grounded in socioscientific inquiry (SSI), and develop a supporting platform that made live sensor data legible and meaningful for learners without a technical background.

Critically, the platform had to be instructor-agnostic — teachers with varying levels of technical comfort needed to facilitate the same inquiry arc without friction.

02 · Process

From sensor data to scaffolded inquiry.

An iterative design process across five phases — each informed by SSI pedagogy and the practical realities of K–12 classroom deployment. Click each step to expand.

Step 01
Needs Analysis
Reviewed SSI and knowledge-building literature. Mapped learner gaps in civic data reasoning — identifying that students lacked both interpretive scaffolding and a sense of personal relevance when engaging with environmental datasets.
Step 02
Curriculum Design
Designed a 6-module inquiry arc structured as: observe → question → investigate → annotate → argue → report. Each module mapped to specific platform features, with differentiation paths for Intro / Intermediate / Advanced learners.
Step 03
Platform Development
Led the tech design sub-team and built the full-stack platform (React / Express / Postgres): sensor data pipeline, live heat map via OpenAQ integration, role-based access for teachers and students, and the annotation system with audit trail.
Step 04
UX Audit
Conducted a structured audit of 4 interface screens and 3 operational guides across both teacher and student flows. Surfaced 7 issues — 4 high severity, 3 medium — including missing empty states, unclear session creation, and numbers without interpretive context.
Step 05
Instructor Guide Redesign
Rebuilt the Instructor Curriculum Guide as a "living pair" with the platform — not static documentation, but two artifacts that evolve together. Platform changes trigger guide updates; pedagogical needs become feature requests.
03 · Artifacts

Design artifacts produced.

Each artifact was designed as part of an interconnected system. Switch between tabs to explore.

M — 0
Orientation
Sign-in · roles · metric names · live demo
≈ 1 hr · Instructor + Students
M — 1
Session Design
Setup · metadata · hypothesis framing
30–45 min · Instructor-led
M — 2
Data Collection
Per-group uploads · session linkage · gap checks
Async / field · All roles
M — 3
Exploratory Analysis
Heat map · group comparison · prompt cards
1–2 hrs · Student-led
M — 4
Annotations & Ethics
Original vs. annotated · rationale · audit trail
45–60 min · Students
M — 5
Reporting & Sharing
CSV export · cite original vs. annotated
≈ 1 hr · Whole group
Short workshop path: M0 → M3 → M4 stands as a complete SSI experience.
High
No direct data input for students — CSV import only; K–12 students preparing CSV files is unrealistic. Interim manual entry form needed.
High
No empty state on first login — blank charts with zero values read as a broken app, not an onboarding state.
High
Session creation entry point unclear — central to teacher flow but not surfaced in any visible tab.
High
Numbers lack interpretive context — PM2.5 = 18 shown without WHO reference, preventing evidence-based SSI arguments.
Medium
Raw Data defaults to all groups — students must filter across 13 columns to find their own data.
Medium
Join Code UI ambiguous — reads as search or creation; no copy button after code is generated.
Medium
Heat Map MODE toggle unlabeled — unclear whether it switches city vs. school data.

Platform and curriculum guide designed as a living pair — when one changes, the other follows.

Well connected
Annotation ethics → M4 covers precisely
Heat Map → M3 spatial pattern activity
CSV export → M5 reporting workflow
Disconnected (pre-audit)
Sensor upload flow → M2 had no procedure
Session creation UI → screen link missing
Join Code sharing → missing from checklist
04 · Outcome

A deployable system connecting data to civic learning.

  • Delivered a fully deployed web platform with live sensor data integration, accessible without technical setup at haetalkim.github.io/tamguingAIR ↗
  • Designed a complete 6-module K–12 inquiry curriculum grounded in socioscientific reasoning — structured to develop both scientific literacy and civic argumentation.
  • Produced a redesigned Instructor Curriculum Guide reducing implementation friction, enabling teachers with varied technical backgrounds to facilitate the full inquiry arc independently.
  • Conducted a structured UX audit surfacing 7 issues across 4 screens — findings directly shaped guide redesign and platform sprint backlog.
  • Demonstrated end-to-end instructional design ownership: needs analysis → curriculum architecture → platform development → instructor support → iterative redesign.
Designer's Reflection

The most significant learning came from the UX audit midway through the project. What we had designed as a "simple" instructor interface carried hidden assumptions about data literacy. Redesigning the guide forced me to think about scaffolding not just for students, but for teachers — a distinction I now treat as foundational to any learning system design.