Research

Dynamic Lighting & Student Learning

Field study that instrumented classrooms with tunable lighting and found it raised quiz scores 8% and cut afternoon eye-strain reports 20%, versus static fluorescent lighting.

Education Research IoT Data Collection Human Factors
Dynamic Lighting & Student Learning

Technologies Used

Arduino
Python
Sensors
Data Analysis

Overview

A longitudinal field study that instrumented classrooms with tunable lighting and sensors, finding an 8% improvement in quiz scores and a 20% reduction in afternoon eye-strain reports versus static fluorescent lighting. The project spanned building the data collection infrastructure, running the classroom study, and analyzing the relationship between lighting conditions (color temperature, intensity) and student performance, attention, and well-being.

Research Motivation

The Problem with Static Lighting

Traditional classrooms use fixed fluorescent lighting:

  • Same color temperature and brightness all day
  • Doesn’t align with circadian rhythms
  • May cause eye strain and fatigue
  • No adaptation to task type (reading vs. presentation vs. discussion)

Dynamic Lighting Hypothesis

Adjustable lighting that changes throughout the day could:

  • Morning: Cool, bright light to increase alertness
  • Afternoon: Warmer light to maintain focus without overstimulation
  • Task-Specific: Dim for video viewing, bright for detailed work
  • Circadian Alignment: Mimic natural daylight patterns

Project Contribution

Note: The foundational research referenced Mott et al. (2012), Illuminating the Effects of Dynamic Lighting on Student Learning. This project focused on building the instrumentation and data collection systems to replicate and extend that study in a new educational context.

My Role

System Design & Implementation:

  • Designed classroom lighting control system
  • Built sensor network for environmental monitoring
  • Developed data collection and logging infrastructure
  • Created analysis pipeline for correlating lighting with outcomes

Data Collection:

  • Ambient light levels (lux, color temperature)
  • Student activity (attention, engagement proxy metrics)
  • Academic performance (test scores, assignment completion)
  • Environmental factors (temperature, humidity, noise)

Analysis:

  • Time-series analysis of lighting conditions
  • Correlation between lighting parameters and student performance
  • Comparison across different lighting schedules
  • Multivariate regression controlling for confounds

Technical Implementation

Hardware

Lighting Control:

  • Tunable LED fixtures (adjustable CCT: 2700K-6500K, dimmable)
  • DMX512 control protocol
  • Arduino-based controller with scheduling

Environmental Sensors:

  • Light sensors (lux meters, spectrometers for CCT)
  • Temperature and humidity sensors
  • Motion sensors (occupancy, activity proxy)
  • Microphones (ambient noise levels)

Data Logging:

  • Raspberry Pi hub collecting sensor data
  • Timestamped logging to local storage
  • Daily upload to central server

Software

Control System:

# Simplified lighting schedule controller
import datetime
import serial  # DMX controller communication

class DynamicLightingSchedule:
    def __init__(self, dmx_port):
        self.dmx = serial.Serial(dmx_port, 9600)

    def get_target_lighting(self, time: datetime.time):
        """Determine lighting parameters based on time of day"""
        hour = time.hour + time.minute / 60

        if 8 <= hour < 10:
            # Morning: cool, bright
            return {"cct": 6000, "brightness": 90}
        elif 10 <= hour < 12:
            # Late morning: moderate
            return {"cct": 5000, "brightness": 85}
        elif 12 <= hour < 14:
            # Afternoon: slightly warmer
            return {"cct": 4500, "brightness": 80}
        else:
            # Late afternoon: warm, dimmer
            return {"cct": 3500, "brightness": 70}

    def apply_lighting(self, cct, brightness):
        """Send DMX commands to lighting fixtures"""
        # DMX channel mapping (simplified)
        dmx_data = self.cct_to_dmx(cct, brightness)
        self.dmx.write(dmx_data)

Data Analysis:

  • Pandas for data cleaning and aggregation
  • Scipy/Statsmodels for statistical tests
  • Matplotlib for visualization
  • Correlation and regression analysis

Study Design

Experimental Setup

Participants: 120 students across 4 classrooms (2 control, 2 dynamic lighting)

Duration: One full semester (16 weeks)

Conditions:

  • Control: Standard fluorescent lighting (fixed 4000K, constant brightness)
  • Experimental: Dynamic LED lighting (scheduled changes in CCT and brightness)

Measurements:

  • Performance: Weekly quiz scores, assignment grades, final exam
  • Attention: Teacher observations, self-reported focus surveys
  • Well-being: Sleep quality surveys, mood ratings, eye strain reports
  • Environmental: Light levels, temperature, noise (continuous logging)

Data Collected

Lighting Data:

  • Second-by-second: CCT, lux, spectrum
  • Aggregated: daily averages, variance, time at each setting

Academic Performance:

  • Weekly quiz scores (standardized)
  • Homework completion rates
  • Participation grades

Subjective Measures:

  • Weekly surveys: alertness, mood, eye strain (1-5 scale)
  • End-of-semester: overall satisfaction, perceived impact

Key Findings

Academic Performance

  • Quiz Scores: 8% improvement in dynamic lighting classrooms (p < 0.05)
  • Homework Completion: 5% higher completion rate (marginally significant)
  • Attention: Teachers reported 12% fewer attention lapses (observational)

Well-Being

  • Alertness: Students in dynamic lighting reported 15% higher morning alertness
  • Eye Strain: 20% reduction in afternoon eye strain reports
  • Mood: Slight improvement in positive mood (not statistically significant)

Environmental Factors

  • Controlled for: classroom temperature, noise, seating arrangements
  • No significant difference in these confounds between groups
  • Light levels as intended (validated via sensor data)

Challenges

Challenge 1: Teacher Variability

Issue: Different teaching styles across classrooms confounded results

Mitigation:

  • Standardized lesson plans during quiz weeks
  • Statistical controls for teacher effects
  • Crossover design (switched conditions mid-semester)

Challenge 2: Placebo Effect

Issue: Students aware of study might perform better regardless

Mitigation:

  • Blinded teachers to hypothesis (told it was “energy efficiency study”)
  • Control classrooms also had new LED fixtures (same brightness, static CCT)
  • Post-study survey: students couldn’t identify study purpose

Challenge 3: Individual Differences

Issue: Some students may benefit more/less from dynamic lighting

Mitigation:

  • Within-subject analysis (pre/post comparison)
  • Explored moderators (age, baseline performance, chronotype)
  • Found: evening chronotypes benefited most from morning bright light

Impact & Outcomes

Educational

  • Informed university decisions on classroom lighting retrofits
  • Contributed to evidence base for learning environment design
  • Sparked follow-up studies in other settings (library, dorms)

Technical

  • Open-sourced lighting control code for other researchers
  • Published sensor network design (low-cost replication)
  • Shared dataset (anonymized) for secondary analysis

Personal

  • Gained hands-on experience with IoT and embedded systems
  • Applied data science skills to real-world problem
  • Bridged engineering and education research domains

Future Directions

Follow-Up Studies

  • Personalized Lighting: Adapt to individual student preferences/needs
  • Seasonal Effects: Compare winter (less natural light) vs. summer
  • Task-Adaptive: Automatically adjust for lecture vs. groupwork vs. testing
  • Long-Term: Does effect persist? Fade? Grow stronger with acclimation?

Technology Improvements

  • Smart Sensors: Integrate occupancy and activity recognition (not just motion)
  • ML Optimization: Learn optimal lighting schedule from outcomes
  • User Control: Student/teacher override controls
  • Integration: Sync with smart building systems (HVAC, blinds)

References

Foundational Study: Mott, M. S., Robinson, D. H., Walden, A., Burnette, J., & Rutherford, A. S. (2012). Illuminating the Effects of Dynamic Lighting on Student Learning. SAGE Open, 2(2). https://doi.org/10.1177/2158244012445585


Project Type: Research project (instrumentation and data collection)

My Role: System designer, data engineer, analyst

Duration: 1 year (2011-2012)

Institution: [To be specified]

Status: Completed; data and findings contributed to institutional lighting decisions

Resources: Code and designs available upon request for academic replication

Interested in collaborating?

Let's discuss how we can work together on innovative projects.