Kensuke Takii's metaphysics paper defines LA's identity with 8 agents
15 years of Learning Analytics finally get an internal ontological grounding.
Kensuke Takii's paper tackles a missing piece in the Learning Analytics (LA) field: a metaphysics derived from its own principles. For 15 years since the first LAK conference, epistemological and ethical debates have flourished, but metaphysical discussions about what LA fundamentally is have been sparse. Takii argues that to answer 'What is LA?' we must look inward, not impose external frameworks. The paper examines what kind of existence the data LA operates on constitutes and identifies eight agents—including learners, educators, and algorithms—as ontological prerequisites. It also clarifies, via the is/ought problem from philosophy, that LA does not derive norms solely from data analysis.
A key contribution is the identification of 'norm-embedded LA,' a class of practices that conflate LA's purpose (e.g., improving learning outcomes) with its operations (data collection and inference). This creates an ontological tension with the field's first principles. The metaphysics outlined here is not an external imposition but surfaces what LA has always implicitly presupposed. Takii discusses connections with related fields like human-computer interaction and acknowledges limitations. The paper offers a rigorous foundation for LA practitioners and researchers to reconsider the field's identity and avoid philosophical pitfalls in their data-driven work.
- Identifies eight ontological agents (learners, educators, algorithms, etc.) as prerequisites for LA.
- Uses the is/ought problem to argue LA cannot derive norms from data alone.
- Coins 'norm-embedded LA' as a problematic practice that conflates purpose with operations, creating ontological tension.
Why It Matters
Provides a much-needed internal philosophical foundation for Learning Analytics, helping practitioners avoid conceptual mistakes in data-driven education.