UNT AI in Action · Dr. Scott J. Warren · University of North Texas
Coherense
Gamified Human-AI Collaboration for Executive Digital Transformation, Training, and Decision-support
Department of Learning Technologies · College of Information
Why digital transformations fail
60–70%
Overall DT failure rate (Gartner, 2025)
42%
Organizational factors (Syed et al., 2023)
38%
Cultural resistance
35%
Leadership gaps
31%
Implementation problems
The core problem is usually not the shiny tool. It is the messy human system wrapped around it.
Syed et al. (2023). International Journal of Information Management. | Oludapo, Carroll, & Helfert (2024). Information Systems Management. | Gartner (2025).
What the app actually does
1
Users move through short mission-based scenarios that simulate real transformation constraints: stakeholder conflict, capability gaps, governance pressure, and uncertain technology fit. (Brown et al., 1989; Collins et al., 1989)
2
Instead of asking the AI to “solve it,” the user is prompted to justify choices, compare options, and explain consequences. (Zimmerman, 2000)
3
The system produces a structured output reviewable for coherence, defensibility, and organizational fit. (Schmarzo, 2020; Warren et al., 2025)
The Products: Coherense and Meridian
Coherense
AI-guided training platform
Executive digital transformation decisions
- Mission-based scenarios with real DT constraints: stakeholder conflict, capability gaps, governance pressure
- Human-AI collaboration: AI reasons alongside the user, not instead of them
- Decision artifacts: every session produces a usable output for real planning
- Three learning arcs: AI Integration, Quantum Readiness, Emerging Tech (ECET)
- Structured feedback on reasoning quality and coherence of choice
Meridian
Narrative learning game
Digital transformation leadership
- Story-driven game: players navigate a fictional organization undergoing digital transformation
- Consequential decisions: choices ripple through stakeholder relationships, resources, and outcomes
- Companion to Coherense: transfers analytical frameworks into applied narrative practice
- Replay and branching: learners test alternate paths and compare outcomes (Warren & Jones, 2017)
- Under ongoing Delphi study validation with executive participants
Games as decision-support tools: a 20-year research lineage
The core executive challenge Business leaders routinely make high-stakes decisions with incomplete, ambiguous, and competing information. Improving that capacity is not a training problem that lectures solve. It requires repeated practice making and defending consequential choices under realistic constraint. (Warren & Jones, 2017)
Research lineage This work extends a 20-year program of game-based problem-solving and decision-support research: from Quest Atlantis (2005) to Opening the Door (Warren et al., 2011) to Broken Window transmedia games (Warren, 2019) to the current platform. Each generation trained critical thinking and decision-making capacity through consequential play.
What Coherense and Meridian add Both platforms are designed as decision-support systems, not content delivery systems. Scenarios surface the assumptions executives bring to technology adoption and force explicit reasoning about tradeoffs before commitment. Feedback develops the metacognitive habit of asking: what do I actually know, what am I assuming, and what would change my choice?
Decision artifacts Every session ends with a structured output the executive can use: an FMEA risk score, SDTDF readiness report, or governance dashboard. The game mechanic and the professional deliverable are the same object, ensuring transfer of problem-solving practice directly into organizational decisions. (Schmarzo, 2020; Warren et al., 2025)
Warren et al. (2011). Computers & Education. | Warren (2019). Journal of Interactive Learning Research. | Warren & Jones (2017). | Warren et al. (2021). | Schmarzo (2020).
Pedagogy: problem-solving and deciding with incomplete information
Decisions with incomplete information
Every scenario presents a realistic dilemma where the executive has enough information to act but not enough to be certain. This mirrors the actual conditions of technology adoption decisions, where waiting for certainty is itself a consequential choice. (Watson & Fang, 2012; Warren & Jones, 2017)
Critical thinking through justification
The platform does not ask what the learner chose. It asks them to justify why. Explaining a choice under competing constraints is the core cognitive act that builds transferable critical thinking capacity, grounded in cognitive apprenticeship theory. (Collins et al., 1989; Brown et al., 1989)
Posture and metacognition
Decision patterns aggregate into a Speed / Governance / Caution posture profile. Surfacing a decision-making style the learner may not recognize in themselves is a precondition for improving it. (Black & Wiliam, 1998; Zimmerman, 2000)
Progressive disclosure as problem structure
Content unlocks in layers: Koan framing, Teaching, Context, then Scenario. The learner must engage with the problem frame before the decision is available. Real decisions arrive with incomplete context, and learning to gather information before acting is itself a skill. (Zimmerman, 2000)
Narrative simulation (Meridian)
Meridian places the executive inside a fictional organization across 20 quarterly decisions. Choices compound: a governance shortcut in Q1 creates a stakeholder crisis in Q7. Developed from the Anytown and Door game lineage, this mechanic trains causal reasoning across time horizons that single-scenario formats cannot. (Warren et al., 2011)
Role differentiation and transfer
Four executive role tracks and a secondary lens (ethics, risk, ROI, governance, readiness) ensure the problem-solving practice matches the decision contexts each participant actually faces, maximizing transfer to organizational practice. (Tomlinson, 2001; Collins et al., 1989)
How it was built: AI-assisted engineering
Spec-driven AI prototyping A 288-line engineering specification (v40, February 2026) was authored first, then provided to Claude (Anthropic, 2024) as a complete prompt-context to generate the full prototype codebase - HTML, CSS, and JavaScript - with zero external libraries.
Architecture Static front-end deployed on a Hostinger VPS. No frameworks or backend required for basic use. Phase 2 adds a Flask + SQLite API for session logging and research data export.
Research instrumentation The spec includes a full behavioral logging schema - panel-open sequences, time-on-task, scenario choices, posture tallies, and tool uptake - enabling pre/post posture-shift measurement as a within-subject research instrument.
Iterative versioning 40 build versions across January–February 2026. Each version updated via revised specification prompts to Claude, demonstrating a human-in-the-loop model where the researcher retains full authorship of requirements, content, and design decisions.
Human-AI collaboration model
The AI layer acts as a reasoning collaborator rather than a generic chatbot (Warren & Beck, 2023)
It helps clarify intent, surface assumptions, and compares strategic options (Warren et al., 2023)
Players identify second-order effects before implementation decisions are finalized (Grotewold et al., 2024)
This ensures leaders know to ask why, can, and should before IT adoption
Warren & Beck (2023). In Spector et al. (Eds.), Learning, Design, and Technology. Springer Nature. | Warren, Beck, & McGuffin K. (2023). In Moore & Dousay (Eds.), Applied Ethics for Instructional Design and Technology. EdTechBooks. | Churchill, Warren, & Grotewold (2024). Decision Sciences Journal of Innovative Education.
Research and validation path
Current study
Delphi Study with Executive Participants
Focus areas
Usability · Learning experience · Guided Human-AI decision-making
There is an ongoing Delphi study with executives examining usability, learning experience, and perceptions of guided human-AI decision-making using training from Coherense and the associated Meridian learning game.
That matters because the platform is not just software - it is a training intervention that needs evidence of learning value and decision quality improvement.
Takeaway: If it changes behavior in organizations, not just screens, it has value.
References
Anthropic. (2024). Claude [Large language model]. https://www.anthropic.com
Black, P., & Wiliam, D. (1998). Assessment and classroom learning. Assessment in Education, 5(1), 7–74.
Brown, J. S., Collins, A., & Duguid, P. (1989). Situated cognition and the culture of learning. Educational Researcher, 18(1), 32–42.
Checkland, P. (1981). Systems thinking, systems practice. Wiley.
Collins, A., Brown, J. S., & Newman, S. E. (1989). Cognitive apprenticeship. In L. B. Resnick (Ed.), Knowing, learning, and instruction (pp. 453–494). Erlbaum.
Gartner. (2025). Top strategic technology trends. Gartner Research.
Churchill, C., Warren, S. J., & Grotewold, K. (2024). Changes to business faculty perceived skills with online teaching tools and educational practices: The pandemic effect. Decision Sciences Journal of Innovative Education, 1-16.
Oludapo, O., Carroll, N., & Helfert, M. (2024). Digital transformation failure factors. Information Systems Management.
Schmarzo, B. (2020). The economics of data, analytics, and digital transformation. Packt.
Syed, R., et al. (2023). Why digital transformations fail. International Journal of Information Management.
Tomlinson, C. A. (2001). How to differentiate instruction in mixed-ability classrooms (2nd ed.). ASCD.
Warren, S. J., Dondlinger, M. J., McLeod, J., & Bigenho, C. (2011). Opening the door: An evaluation of the efficacy of a problem-based learning game. Computers & Education, 58(1), 397-412.
Warren, S. J., Beck, D. E., Najmi, A., & Darby, D. (2019). Transmedia play to teach computer literacy, global thinking, and rudimentary instructional design: Instructors reflect on teaching with Broken Window. Journal of Interactive Learning Research, 30(4), 505-537.
Watson, W. R., & Fang, J. (2012). PBL as a framework for implementing video games in the classroom. International Journal of Game-Based Learning, 2(1), 77-89.
Warren, S. J., & Beck, D. E. (2023). The Ethical Choices with Educational Technology Framework. In M. J. Spector, B. B. Lockee, & M. D. Childress (Eds.), Learning, Design, and Technology. Springer Nature.
Warren, S. J., Beck, D., & McGuffin, K. (2023). In support of ethical instructional design. In Moore & Dousay (Eds.), Applied Ethics for Instructional Design and Technology (pp. 41-62). EdTechBooks.org.
Warren, S. J., & Jones, G. (2017). Learning Games: The Science and Art of Development. (D. Ifenthaler & D. Esreyel, Eds.). Springer.
Warren, S. J., Roy, M., & Robinson, H. (2021). Business simulation games: Three cases from supply chain management, marketing, and business strategy. In D. Ifenthaler (Ed.), Game-based learning across the disciplines. Springer.
Warren, S. J., et al. (2025). Dynamic Systems Engineering for digital transformation. Manuscript under review.
Zimmerman, B. J. (2000). Attaining self-regulation. In Boekaerts et al. (Eds.), Handbook of self-regulation (pp. 13–39). Academic Press.
← → keys