AACSB accreditation
Every score traces to an explicit rule — auditable, mappable to learning outcomes, and reportable as Assurance-of-Learning data.
The evidence behind vLeader
vLeader runs on the KRY engine — a symbolic-AI system of hand-coded causal rules, built in 2002 and deliberately never replaced by a language model. This page sets out why that choice matters, what our own outcome data shows, the method used to produce it, its limits, and every scholarly work published about the simulation with its design and its limits stated alongside it.
The KRY engine
Symbolic AI where you would expect a language model
The simulation is driven by hand-coded microstrategies organized into five categories — Power, Ideas, Tension, Work, and Relationship. Each is a small, interpretable causal rule. NPC behavior is selected by a transparent utility calculation over those rules. No model generates it.
Its lineage is classic interactive-AI and agent-based modeling: Mateas & Stern's Façade (2003), Schelling's game theory, Epstein–Axtell agent-based modeling, and the behavior-tree / utility-AI tradition. It is GOFAI — “good old-fashioned AI” — running in production in 2026.
Because there is no language model, the system is rule-governed, replayable, debuggable, and hallucination-free. The rules are fixed and hand-coded — no model generates the consequence, and every score maps to a stated rule, for every student, in every cohort, every semester. Exact dialogue lines vary between runs; the causal logic does not.
// the differentiator, in one line
Fixed causal rules.
0 large language models.
0 hallucinations.
// same rules, same scoring — every cohort
Why fixed rules matter in 2026
Auditable, rule-traceable outcomes every time
Government & defense
No external API calls, no model drift, no hallucination risk — the same scenarios and scoring rules, every cohort. Defense Acquisition University is a current adopter.
Healthcare & legal
Scenarios governed by fixed, inspectable rules that do not change between runs — the property regulated-domain training and review actually require.
Longitudinal research
Compare cohorts across years against the same scenarios and the same scoring rules — the rulebook does not drift between semesters.
Outcome data
What our own data shows
Reflection scores, summer 2026
Paired pre/post reflections · Model-graded rubric · 15 June – 28 August 2026
Students complete a graded reflection before and after the debrief. Across the paired reflections from that term, 62% scored higher afterwards, and improvement outnumbered decline better than two to one — 61.9% up against 28.1% down. At a three-point threshold on the ten-point rubric the ratio holds: 27.6% against 13.3%.
The obvious objection is regression to the mean, so here is the test that rules it out. Among students who had already scored 9 out of 10 before the debrief — with almost no headroom left — 63% still improved. A pure regression artifact would have pushed that group down instead.
At student level: 58% show net improvement across their reflections, and about one in eight averages nearly four points of gain. It does not land the same way for everyone, and we would rather publish that than an average that hides it.
Single-group before-and-after with no control group: this shows that scores move; whether vLeader caused the movement against an alternative is beyond what it can establish. A model grades the rubric, with no human rater involved, and pre/post scores correlate at r = 0.44. It covers one summer term, and only sections whose instructor uses the reflection loop appear, so the students in it are a self-selected minority of those enrolled in the period.
Independent literature
Everything published about this simulation
Below is every scholarly work we can confirm is about vLeader. We checked each one against the publisher’s own record, read the full text wherever it is open, and wrote the design limitation into the same entry as the finding. Where an author has a relationship with the publisher of the software, the entry says so. None of these is a randomized controlled trial.
-
01 Open access
Dalla Rosa, A. & Vianello, M. (2015)
Procedia — Social and Behavioral Sciences 171, 1065–1074 · ICEEPSY 2014 · doi:10.1016/j.sbspro.2015.01.267
“On the Effectiveness of a Simulated Learning Environment.” Forty-five students at the University of Padua, all female, used vLeader as the main tool of a month-long experiential workshop inside a distance-learning human-resources course. The Multifactor Leadership Questionnaire was administered before and after. The result needs reading carefully. Across the whole sample the authors found that examination of the data “revealed only marginal changes from pre-test to post-test.” Their significant finding is conditional on the student: dynamism interacted with the pre/post observation on the MLQ outcome scale, F(1, 42) = 10.45, p = 0.002, ηp² = 0.20, and the gain concentrated among students who scored low on dynamism.
One intact cohort, self-reported measures, no randomization. The authors state they “did not have the chance to assess changes in a control group,” flag the small and entirely female sample, report weak internal consistency on some MLQ scales, and call for multi-source studies on bigger samples with a control group before the results are treated as definitive.
-
02 Open access
Putman, P. G. (2012)
Doctoral dissertation, Cleveland State University · ETD Archive 249 · ERIC ED545735
“Virtual Simulation in Leadership Development Training; the Impact of Learning Styles and Conflict Management Tactics on Adult Learner Performance.” The largest sample in this literature — 349 undergraduate and graduate students drawn from 16 courses at four US universities in Ohio, Kentucky and Maryland, of whom 302 completed simulation modules. Putman measured conflict-management tactics with the POINTS scale and learning style with the Kolb Learning Style Inventory, then tested how far either predicted the Power, Ideas, Leadership and Overall scores the simulation itself produces. Individual influence tactics predicted the Power score at small effect sizes, with reported correlations around r = .13 to .19 and r² under .04.
Correlational and single-group, with no control group and no randomization. Two of the four institutions were referred to the researcher by SimuLearn, which belongs in any reading of how the sample was recruited. The dissertation predicts in-simulation scores from prior traits. It establishes nothing about whether using vLeader changed a student.
-
03 Paywalled
Ruskov, M. & Seager, W. (2011)
Serious Games Development and Applications (LNCS 6944), 49–60 · SGDA 2011 · doi:10.1007/978-3-642-23834-5_5
“What Can Bits Teach about Leadership: A Study of the Application of Variation Theory in Serious Games.” The only design in this list that used a control group. Ruskov and Seager followed the learning-study method and, in the words of their abstract, “introduced experimental and control groups in order to allow for more direct comparison.” They state the constraint they worked under just as plainly: “we did not have control over the employed serious game (vLeader) and thus had to implement our experiment through its supplementary materials.” Their own reading of the result is that the study “provides weak evidence of greater stakeholder awareness by the group that has experienced variation.”
The intervention landed on the supplementary materials around the simulation, so the comparison runs between two ways of teaching with vLeader. The authors call their evidence weak and their findings exploratory. The abstract states no sample size, and the full text is behind a subscription we have not read.
-
04 Open access
Gurley, K. & Wilson, D. (2011)
Journal of Instructional Pedagogies, vol. 5 · ERIC EJ1096964 · Fayetteville State University (an HBCU)
“Developing leadership skills in a virtual simulation: coaching the affiliative style leader.” Thirty-seven MBA students across three sections of an Organizational Behavior class completed the HayGroup Managerial Style Assessment, then worked through the first three simulation scenarios over six weeks. The authors compared affiliative-style students against all other styles and report that affiliative-style students scored lower on developing and using power, and that scores improved after repeated attempts at the first scenario.
An observational comparison between style groups. The study used no control group and no randomization. It is a single study, from 2011, in a practitioner journal without an impact factor. We cite it because it exists and is real. It does not settle the question.
-
05 Open access
Young, A. M. (2012)
International Journal of Business and Social Science 3(19) · California State University, Los Angeles
“Assessing Leadership Skill Development: Implementing vLeader© Simulation Software to Capture Advances in Leadership Skills among Undergraduate Students.” Forty-eight undergraduate business students used vLeader across two intact classes — 21 with no prior instruction in leadership and management, and 27 who had completed a full course in it. Young reports that first attempts scored about the same in both groups and that improvement over repeated play was generally larger among the students with prior instruction.
Two intact classes, no randomization, no control group. Weigh the venue as well as the finding: at the time of publication the journal was published by the Center for Promoting Ideas, which appears on Beall’s list of potentially predatory publishers. We list the paper because it names the product and states its method. The venue supplies no editorial guarantee.
-
06 Paywalled
Standifer, R. L., Thiault, P. & Pin, R. (2010)
Journal of Leadership & Organizational Studies 17(2), 167–176 · doi:10.1177/1548051810366714
“Leadership Development in an Electronic Frontier: Connecting Theory to Experiential Software Through Supplemental Materials.” A companion text written so that students can connect what happens inside the meeting to leadership theory taught across a range of courses. The abstract names the product directly — “a simulation-based product, SimuLearn’s vLeader.” We have read the abstract and the publisher record. The full text sits behind a subscription and we have not read it.
A pedagogy paper. It reports no outcome measure, no sample and no comparison. Until August 2026 this page attributed the work to “Goosen & Van Tonder (c. 2010).” That attribution was wrong. The authors are Standifer, Thiault and Pin, verified against the SAGE record and the publisher’s own Crossref deposit. Disclosure worth having: Standifer also wrote vLeader Professional: Developing Effective Leaders, published by SimuLearn in 2010.
-
07 Paywalled
Johnson, S. A. & Luo, J. (2012)
Increasing Student Engagement and Retention Using Immersive Interfaces · Cutting-edge Technologies in Higher Education 6C, 161–202 · doi:10.1108/S2044-9968(2012)000006C009
“Engaging Chinese Students and Enhancing Leadership Development through Virtual Simulation: A Cross-Cultural Perspective.” A final-year course at the University of Greenwich Business School was rebuilt as personalized leadership development, and the teaching team used vLeader to draw in Chinese students who might otherwise stay passive in a Western classroom. The abstract names the product and says the chapter “outlines our integrated approach.”
A description of teaching practice. The abstract reports no measurement, no sample and no comparison. The chapter sits behind a subscription and we have read only the abstract.
-
08 Closed access
Berizzi, G., Bonaldo, F., Dalla Rosa, A., Frison, D., Petromilli, A., Serbati, A., Tino, C., Fedeli, M. & Vianello, M. (2017)
Chapter in Developing generic competences outside the university classroom, Editorial Universidad de Granada, 41–57 · ISBN 978-84-338-6016-3 · hdl:2158/1147947
“VLeader© in the Academy: Learning Through an Educational Simulator.” Nine authors from the University of Padova describe running vLeader across two courses, the first from March to May 2014 and the second from October to December 2014. Two of them, Anna Dalla Rosa and Michelangelo Vianello, also wrote the 2015 study at the top of this list.
A programme description. The repository record gives no sample size and no control group, and the chapter itself is closed access, so this entry rests on its catalogue record and we have not read the text.
What we left out, and why
Zoroja, Pejić Bach and Miloloža (2019, MIPRO) survey simulation-game platforms for leadership. Their abstract never names vLeader and the full text is paywalled, so we could not confirm the paper belongs here. Rabelo, Almalki and Gutiérrez (2017, Revista Espacio I+D) propose adapting NASA’s 4-D leadership framework to avatar-based simulation; their full text is open and contains no mention of vLeader or SimuLearn. Database searches for the string “vLeader” also return a 1900 strawberry catalogue, coronavirus transcription-regulatory-sequence work, and DC microgrid voltage control.
What we do not yet know
No randomized controlled trial of vLeader has been published. The strongest design in the list above is a single cohort measured before and after with a validated instrument and no control group, and our own reflection data has the same shape. Every study here can show that scores move. None of them can show what would have happened to the same students under a different method. A committee will ask that question, so we answer it here.
Track record
24 years, deliberately built
-
2002
The KRY symbolic-AI engine is built; Virtual Leader ships — years before the LLM era.
-
2004
Training & Development Journal names it Best Online Training Product of the Year.
-
2011
Gurley & Wilson publish an MBA classroom study at Fayetteville State University.
-
2024
Rebranded as vLeader — same rule-based engine, refined across hundreds of thousands of decisions.
Theoretical foundations
The frameworks your syllabus already cites
Goleman's leadership styles
Directive, visionary, affiliative, democratic, pacesetting, coaching.
Tuckman's team development
Forming, storming, norming, performing — enacted as the meeting unfolds.
Thomas-Kilmann conflict modes
Competing, collaborating, compromising, avoiding, accommodating.
Kolb's experiential cycle
Experience → reflect → conceptualize → experiment — the product's own loop.