Why Competency Is More Than a Score
A score describes observed performance at a particular point in time. Competency is more complex: it can evolve, persist, decay, and behave differently across contexts.
Research & Insights
Understanding learning beyond the score
MPCSM (Multi-Parameter Probabilistic Competency State Model) is a conceptual framework that represents competency as four evolving dimensions: mastery, retention, pressure robustness, and fragility.
Learning systems have become increasingly effective at measuring activity, performance, and outcomes.
But a score captures a moment.
It does not necessarily tell us whether what was learned will persist, whether capability will remain stable over time, or whether performance will hold under changing conditions.
At Edculcate, we explore computational approaches to understanding these deeper dimensions of learning and capability.
Our work sits at the intersection of learning science, behavioral analytics, probabilistic modeling, and scalable technology.
Ideas we're exploring
The Four Dimensions of Competency
Can competency be understood through multiple interacting dimensions rather than a single measure?
Our current research explores four:
Mastery
How strongly capability has been acquired.
Retention / Decay
How persistently that capability remains available over time.
Pressure Robustness
How reliably capability remains stable under changing or demanding conditions.
Fragility
How susceptible competency is to instability, perturbation, and accelerated degradation.
Can Competency Become Fragile?
A learner may appear to perform consistently while underlying competency becomes increasingly susceptible to disruption. This raises a different question from “How well is the learner performing?”
How stable is the capability behind that performance?
From Assessment Scores to Competency Trajectories
Two learners can produce similar assessment results while following very different trajectories over time. Understanding those trajectories may reveal information that isolated assessments cannot.
Can We Predict Competency Failure?
If competency changes continuously, can changes in its trajectory provide early signals of degradation before visible performance failure occurs?
What If We Could Measure Readiness, Not Just Learning?
Learning and readiness are related, but they are not the same. The more difficult question may be whether a person can reliably apply what they know when it matters.
Featured Research
MPCSM
Multi-Parameter Probabilistic Competency State Model
A conceptual computational framework for multidimensional competency estimation within adaptive learning systems.
The Multi-Parameter Probabilistic Competency State Model (MPCSM) proposes that competency should not always be represented as a single score or static state.
Instead, it models competency as a continuously evolving multidimensional probabilistic process.
The current framework represents competency through four independently evolving parameters:
- Mastery
- Retention / Decay
- Pressure Robustness
- Fragility
The model explores how observable learner interactions, reinforcement intervals, contextual conditions, performance variability, and historical behavioral dynamics can contribute to understanding competency trajectories.
Why this matters
A learner with high current performance may not necessarily have stable long-term competency.
Two learners may achieve similar assessment results while differing substantially in:
- long-term retention
- response to reinforcement interruption
- performance under contextual stress
- susceptibility to instability
A multidimensional representation may therefore provide a richer characterization of competency than scalar assessment measures alone.
Read the researchA different way of thinking about competency
Traditional assessment tends to ask:
How well did someone perform?
A dynamic competency perspective asks additional questions:
How stable is that performance?
How long is the capability likely to persist?
How does it behave under different conditions?
Is the underlying competency becoming less stable even before performance visibly declines?
These questions shift the focus from isolated outcomes toward competency states and trajectories.
From research to technology
Research is only one part of Edculcate's work.
We are exploring how computational models of learning and competency can become practical technology for learning intelligence and capability readiness.
The technology platform is currently in stealth.
As the work develops, this section will evolve from research concepts toward the systems, methods, and technologies that emerge from them.
Explore our technologyQuestions
What is MPCSM?
MPCSM (Multi-Parameter Probabilistic Competency State Model) is a conceptual framework that represents competency as four evolving dimensions: mastery, retention, pressure robustness, and fragility. It is a 2026 conceptual research framework by Amit Basu at Edculcate India Pvt. Ltd. A patent has been filed; formal publications are forthcoming.
What are the four dimensions of competency?
The four dimensions are mastery, retention, pressure robustness, and fragility. Together they describe whether capability is present, whether it lasts, whether it holds under stress, and how easily it can collapse. MPCSM treats these as evolving parameters rather than a single score.
How is competency different from a score?
A score captures performance at a moment; competency is a changing state that can persist, decay, or fail under pressure. Edculcate models that state across mastery, retention, pressure robustness, and fragility. Scores remain useful observations; they are not the whole of capability.
Is Edculcate's product available?
No. Edculcate's technology platform is currently in stealth. The company is translating research into product, but the product is not available for purchase or public use.
Research status
MPCSM — Multi-Parameter Probabilistic Competency State Model
- Author
- Amit Basu
- Organization
- Edculcate India Pvt. Ltd.
- Year
- 2026
- Status
- Conceptual Research Framework
- Intellectual Property
- Patent filed.
Formal publications, preprints, and subsequent research will be added here as they become available.
View publication / preprint — forthcoming