STEM-learning and engagement in school making activities: Linking lesson design, psychological distance, sensory input, sympathetic activation, learning, and engagement

About the project

Why Do Students Disengage from STEM?

Many students lose interest in science, technology, engineering, and mathematics (STEM) during secondary school, which can lead to dropping out or avoiding STEM careers. Traditional teaching often relies on abstract symbols and textbooks, which can feel psychologically distant to students. This dissertation investigates whether making activities, where students design and build physical artefacts such as robots, solar catchers, or DNA models using both digital and craft materials, can counteract disengagement and improve learning.

The Challenge of Measuring Engagement

To understand whether making activities actually work, we need to measure student engagement accurately. Self-reports and surveys are useful but capture only retrospective generalisations. This thesis developed and validated a novel method that combines video analysis with physiological sensors. By analysing movement and sound levels in classroom videos and comparing them to electrodermal activity (EDA), a measure of sympathetic nervous system activation collected via wearable sensors, the study confirmed that observable behaviours (movement and sound) significantly predict physiological arousal. This provides researchers with an affordable, objective tool to assess moment-to-moment engagement without interrupting the class.

The Sensory Learning Model: Why Making Works

At the heart of the thesis is the Sensory Learning Model, which bridges neuroscience and educational practice. The model proposes a causal chain: lesson design influences sensory input, which affects psychological distance and physiological activation, ultimately impacting learning and engagement.

Psychological distance refers to how far away or abstract a concept feels. When students work with physical materials they can touch and manipulate, the distance shrinks. This activates multiple senses (multimodality), increases sympathetic arousal towards optimal levels, and enhances memory formation through perceptual learning pathways.

The 3M Model: Practical Design Guidelines

Through four iterative studies (Design-Based Research) including one randomised controlled trial (RCT) and two quasi-experiments, the thesis tested specific lesson design variables and distilled them into the Multi-Modal Maker Model (3M Model). The overarching key findings for teachers are:

• Pre-activity STEM concept teaching: Teach the relevant STEM concepts before students start building. When students lack this foundation, making becomes craft without learning.
• Semi-open materials: Avoid rigid step-by-step instructions (too constraining) and avoid completely open projects (too difficult). Instead, use materials that provide structure, such as seven phases (research, plan, make, test, evaluate, improve, report) combined with reflection questions.
• Materials table: Provide a table with abundant materials for students to choose from rather than giving each group an identical kit. This increases ownership and reduces psychological distance.

Remote education: Avoid digital teaching of making activities when in-person education is possible. This utilises the engagement and learning benefits of making activities to a larger degree.

These principles were validated across different settings, including remote education during COVID-19 (where the digital format increased psychological distance and reduced engagement) and a large-scale study with 509 general education students, showing that the guidelines work beyond just STEM-interested youth.

Individual Differences: The STEM Scale

Not all students benefit equally from the same activities. The thesis developed a STEM scale that combines spatial ability (the capacity to mentally manipulate objects) with engagement levels to identify students who need differentiated support. Spatial ability is a strong predictor of STEM success, and making activities appear particularly beneficial for students with high spatial ability but lower academic engagement, offering them a pathway back into learning.

Conclusion: A Bridge Between Brain and Classroom

This dissertation demonstrates that how you design hands-on activities matters just as much as whether you include them. By reducing psychological distance, making STEM concepts tangible, concrete, and immediate, teachers can optimise students' physiological activation and attention, leading to deeper learning. The Sensory Learning Model provides a theoretical foundation for why hands-on, multimodal learning works, while the 3M model offers concrete, evidence-based guidelines for lesson planning. For policymakers and educators seeking to reduce dropout and enhance STEM learning, the message is clear: let students build, but build the learning design carefully.

Background

• The duration of the PhD-project is four years, and is due to end in March 2026
• Main supervisor: Imac Maria Zambrana, ONH
• Supervisor: Thormod Idsøe, Department of special needs education, UiO

Financing

This work is supported by the Norwegian Research Council under grant number 308895 and Akershus County

(All photos by Ellen Egeland Flø)