AIEd Knowledge BaseAIEd KB
On this page

Concept

Theories and Frameworks

created 2026-09-19 · updated 2026-09-19

Theories and frameworks — the map of the explanatory and organizational structures this knowledge base uses. Theories explain why learning happens (Learning Theories, Self-Determination Theory, Sociocultural Learning, activity theory); frameworks organize design, teaching, and adoption decisions (Technological Pedagogical Content Knowledge (TPACK), SAMR, technology adoption models, ICAP, universal design for learning); measurement models make learning claims testable (item response theory, Assessment Validity, Self-Report Measures). Use this page when you want to know which lens a finding rests on — and what that lens can and cannot support.

Questions to Consider

  • A theory predicts what happens when something changes; a framework tells you what to attend to. When a study reports a gain under a framework's banner but makes no prediction that could have failed, what has actually been learned?
  • TPACK, SAMR, UDL, the adoption models and ICAP all predate large language models. Which parts of them still hold when the tool can write, explain, and adapt on its own — and how would you tell?
  • Adoption is often reported as movement through stages. If a colleague reports that their course moved "from substitution to redefinition" this year, what independent evidence would you ask for before believing it?
  • A measurement model can be fitted, compared, and shown to be wrong; a framework cannot. When you decide whether to trust a claim about AI and learning, where does your confidence actually come from — the framework that named the outcome, or the instrument that measured it?
  • Much of this field applies borrowed theory rather than testing it. Pick a finding you rely on: could it be reframed as a test of a theory, and what measurement would that require?
  • Philosophy asks whether a goal was worth pursuing; theory asks whether the mechanism behind it is real. Which of those two questions does your institution's AI strategy leave unanswered?

Theory, framework, model: what each word promises

The three labels are used loosely in the literature, but they promise different things, and the difference decides what a claim built on them can support:

  • A theory explains a mechanism. It names the parts that do the work and predicts what happens when they change: self-determination theory predicts that satisfying autonomy, competence, and relatedness raises motivation, and it fails if it does not; sociocultural theory treats learning as mediated by tools and social interaction; activity theory takes the whole activity system, contradictions included, as the unit of analysis. Theories are falsifiable, which is why studies that test them can report effect sizes.
  • A framework organizes or prescribes. It names the components to attend to and how they relate, usually without predicting magnitudes: TPACK names the knowledge a teacher blends; SAMR stages substitution through redefinition; ICAP ranks engagement modes as passive, active, constructive, interactive; UDL prescribes multiple means of representation, action, and engagement; technology adoption models frame adoption as intention driven by perceived usefulness and ease.
  • A model is a formal representation. In this knowledge base that usually means measurement or computational modeling — item response theory for item and ability estimation, knowledge tracing and learner modeling for estimating a learner's state over time. A model can be fitted, compared, and shown to be wrong in a way a framework cannot.
  • The labels overlap, and that is fine. Community of inquiry is both a theory of the learning experience and a design framework; activity theory is explanatory and analytic at once. Read the label as a clue about what the source claims, not as a filing category.

Where each node lives

The knowledge base files these nodes by topic, so an inventory is spread across sections:

How this differs from philosophy of AI in education

Philosophy of AI in education asks normative and conceptual questions: what education is for, what counts as a good learner, whether a machine can teach, and what the word "intelligence" in the field's name commits us to. It interrogates the aims and categories the other two levels take for granted, and it does not predict effect sizes. The distinction matters in practice: a study can be philosophically naive and theoretically sound, or philosophically rich and empirically empty. Use philosophy to ask whether the goal was worth pursuing; use theories to ask whether the mechanism is real; use frameworks to ask whether the design attended to the right things.

How this differs from theory development in AI in education

Theory development is the meta-activity — how the field builds, borrows, adapts, and tests theory: which constructs it imports from psychology and the learning sciences, which it coins itself, how it theorizes about a technical target that changes every year, and the standing critique that much AIED work applies existing theory rather than testing it. This page is the inventory of what the field currently holds; theory development is about producing and revising the inventory.

Finding the right lens

  • Instructors starting from a problem rather than a theory can work backwards: if learners are not engaging, ICAP and SDT name different causes and suggest different fixes; if the question is whether to adopt a tool at all, adoption models and SAMR ask different things about it.
  • Learning designers get the most from the design frameworks — UDL, Learning Design, and Scaffolding — paired with a theory that predicts what will happen when the scaffolding is removed.
  • Researchers should state which node the study claims and whether the design can actually test it; measurement models (IRT, Assessment Validity, Self-Report Measures) decide whether the reported outcome supports that claim.
  • Administrators meet frameworks as adoption and change questions — adoption models, Change Management, SAMR — where a stage model is often used as a maturity story rather than an instrument.

What frameworks cannot do

Frameworks are not evidence. They are borrowed, usually from pre-LLM contexts, and localized by whoever applies them; they can be used as branding; and stage models invite checkbox adoption that reports movement through levels rather than learning. Claims that rest on a framework should be read alongside the field's cross-cutting limitations, the validity of whatever measured the outcome, and the known limits of self-report.

Connected Concepts

Connected Articles

Metadata

Learning and instructionLearning Theories
Assessment and measurementAssessment Validity, Item Response Theory
Research methods and evaluationResearch Methods in AIED
Foundations of AI in educationAI in Education, Theory Development in AI in Education, Philosophy of AI in Education
SourceHistoryEditChat

Embed this page

Copy the code below to embed a chromeless version of this page in a learning management system or other website. The embedded view hides the site header, navigation, and footer.