Research, Pillar 03
Education & the Next Engineer
How software engineers learn has received far less empirical attention than how they work. This research is an attempt to close that gap.
A recurring question across this research programme is what the field actually knows about engineering education, not what educators assume works, but what controlled and longitudinal studies find. The work spans the full educational pipeline: secondary school, university, and the professional communities where practitioners continue to develop throughout their careers. The consistent finding across all levels is that active, participatory formats outperform passive ones, and that human connection is not a nice-to-have but a precondition for learning that lasts.
Computational Thinking and the Early Pipeline
Learning agile software development at secondary school level is not a standard curriculum experiment. A study investigating this in a high school context found that agile practices, when properly adapted, gave younger learners a concrete framework for managing complexity, not just a set of rituals, but a way of structuring work that transferred to other domains.
A related study applied agile practices specifically to the teaching of computational thinking. The findings showed that iterative, team-based formats changed how students engaged with abstract problem-solving. Passive instruction produced surface-level familiarity; the active format produced transferable reasoning skills. Computational thinking is now a core literacy expectation for engineering graduates. The evidence suggests the method of teaching it matters as much as the decision to include it.
University-Level Pedagogy
Test-first programming, writing tests before writing code, is widely recommended but inconsistently taught. A structured assessment of the approach found that when test-first programming is taught with explicit scaffolding and feedback cycles, it changes how students conceptualise software quality from the outset. The effect was not merely procedural: students taught this way reasoned differently about correctness, not just about test coverage.
A study of cooperative thinking examined a new framework for structuring collaborative problem-solving in software engineering education. Cooperative formats produced measurable gains in collective reasoning that individual work did not replicate. The mechanism is consistent with the broader research on psychological safety: when students feel safe to reason out loud together, the quality of their thinking improves.
Remote Education and the Pandemic as Natural Experiment
The shift to remote education during the COVID-19 pandemic was unplanned and rapid. A study evaluating remote education strategies during this period found that outcomes were more uneven than the optimistic accounts suggested. The conditions that predicted successful learning were interpersonal, not technological. Platform quality mattered less than instructor presence, structured interaction, and the preservation of community.
This parallels findings from the remote work research in Pillar 02. In both contexts, professional and educational, the quality of human connection proved to be the critical variable. The technology mediates that connection; it does not replace it.
Professional Learning Communities and Liberating Structures
Engineering education does not end at graduation. The conferences and community events where practitioners exchange knowledge are, in effect, the continuing education infrastructure of the field. They are also, empirically, poorly designed for learning.
A study published in Communications of the ACM examined this problem and tested an alternative. Liberating Structures, facilitation techniques including Impromptu Networking, 1-2-4-All, Troika Consulting, and Fishbowl Conversations, were applied at the Copenhagen Symposium and at major SE venues. When participants were structured into active roles rather than passive ones, the quality of cross-disciplinary connection improved, shared problems surfaced faster, and outputs were more durable. The Copenhagen Manifesto emerged from exactly this process.
The implication for the field is structural. If the goal of a conference is knowledge creation rather than knowledge broadcast, the format has to change. The evidence that it can change, and what happens when it does, is now documented.
The Unifying Question
What connects these studies, from high school agile experiments to professional conference design, is a single empirical claim: the way engineers are educated shapes not only what they know but how they think and collaborate. Getting that right is not a peripheral concern. It is the condition under which everything else in this research programme becomes useful.