Research foundations

We didn't invent challenge-based learning. We're the digital expression of an 80-year tradition.

Ewance sits inside a well-developed intellectual tradition: Action Learning, the experiential learning cycle, and the modern science of authentic assessment. The technology is new — AI for matching and feedback, blockchain for portable credentials. The pedagogy is older than the internet.

Published research Modern research Collaborate

Published research

We publish our reasoning in the open.

Before we have multi-cohort outcome data, we'd rather set out our thinking than market loosely. Our first working paper makes the theoretical case behind the platform — a conceptual synthesis, openly licensed, and citable by DOI.

PreprintOpen accessCC BY 4.0

Demonstrated Capability as a Lower-Noise Signal of Job-Readiness

A Conceptual Synthesis of Experiential Learning, Authentic Assessment, and Labour-Market Signalling

Why a completed, provenance-verified piece of work is a lower-noise indicator of job-readiness than a grade — and why, in an age of generative AI, the value sits in verifying authentic authorship, not in the artefact alone.

Author: A. Acosta, PhD · Aurea CV OÜ (Ewance)
Published: 1 June 2026 · Zenodo · v1.0
DOI: 10.5281/zenodo.20496464

Read the paper

The pedagogical anchor

The evidence the paper builds on.

The working paper above synthesises four established bodies of work. In brief — with the primary sources beneath, for anyone who wants to read further.

The pedagogical anchor

Action Learning, since 1940

Reginald Revans (1907–2003), a British physicist who rebuilt UK industrial training after WWII, formalised Action Learning: people learn fastest tackling real, unfamiliar problems alongside peers who ask honest questions. His formula — L = P + Q, learning equals programmed knowledge plus questioning — has been applied across industry, healthcare, and higher education for 80+ years. An Ewance challenge is its digital expression.

Revans Academy — University of Manchester

The empirical record

Experiential and project-based learning work

Kolb's experiential-learning cycle (1984) runs all four of its stages inside a single challenge; the Buck Institute / PBLWorks gold-standard framework maps almost one-for-one onto how challenges are structured; and Barnett & Ceci's transfer-of-learning taxonomy (Psychological Bulletin, 2002) explains why skills practised on synthetic, decontextualised tasks transfer poorly — the case for authentic, domain-grounded work over textbook exercises.

Barnett & Ceci (2002) — Psychological Bulletin

AI as a learning amplifier

AI helps — when wired into the doing

After thirty years of intelligent-tutoring research, the answer is yes: Kulik & Fletcher's 2016 meta-analysis of 50 controlled studies put the median effect at d ≈ 0.66 (roughly the 50th → 75th percentile), and VanLehn (2011) found modern tutoring systems on par with expert one-on-one human tutors. Ewance treats AI as a feedback amplifier — explaining rubric scores, flagging weak reasoning — not as a substitute for the work.

Kulik & Fletcher (2016) — Review of Educational Research

The credential layer

Portable, verifiable proof of the work

What Action Learning lacked for 80 years was a way for the learner to leave with durable, portable proof. The W3C Verifiable Credentials standard and Open Badges 3.0 changed the architecture: a credential can be issued, held, and verified across platforms without the verifier contacting the issuer. Ewance issues in these formats and anchors each credential's hash on-chain via LearnCoin — so the proof outlives the platform.

W3C Verifiable Credentials Data Model 2.0

Sources & further reading

For academic collaborators

Want to study what happens here?

We're actively interested in research partnerships: longitudinal studies, comparative pedagogy, credential adoption in industry, AI-assisted feedback as a learning tool. We'd rather collaborate seriously than market loosely.

Anonymised cohort dataScoped to your study, under a data-use agreement that your ethics committee can review.
Co-authorshipWe'll publish with you, not just for you. Pre-registration welcome; null results encouraged.
IRB cooperationWe'll work with your institutional review board on consent flows and data-handling protocols.
Multi-site studiesReplication across cohorts, programmes, and regions — the kind of design the project-based learning literature has long called for.

Propose a partnership