Asian Journal of Education and Social Studies

Few proposals in educational psychology have travelled as far into classroom practice on as thin an evidential base as the theory of multiple intelligences. Since its educational translation in the late 1980s, the theory has generated a large family of instructional strategies built on the premise that learners possess relatively independent cognitive capacities that teaching can profitably target. Quantitative syntheses of classroom applications report unusually large achievement gains, yet the psychometric and neuroscientific claims on which the strategies rest remain unconfirmed, and a substantial critical literature now classifies the theory as an educational neuromyth. This review examines that contradiction rather than resolving it by assertion. It assembles evidence from theoretical, psychometric, neuroscientific, intervention, implementation and teacher-belief literatures, and evaluates the strength of inference each supports. Three findings structure the synthesis. First, the claim of intelligence independence is not supported by the available factor-analytic work, and no study has measured neural processing during activity characteristic of a nominated intelligence, so alignment analyses cannot function as validity tests. Second, the pedagogical family that travels under the theory's name is internally heterogeneous: strategies that pluralise the representation of content are logically separable from strategies that match instruction to a diagnosed learner profile, and only the latter depend on the contested theoretical claims. Existing effectiveness syntheses do not distinguish these designs, which makes their pooled estimates difficult to interpret. Third, the intervention literature exhibits recurrent design weaknesses, including small samples, inactive control conditions, undocumented instructional dosage and outcome instruments constructed for single studies, that leave the reported effects attributable to well-established mechanisms unrelated to the theory. The educational implications are therefore asymmetric. Practices emphasising varied representation, expanded assessment formats and broadened talent identification retain independent justification; profile-matched instruction does not. Priorities for future work include design-disaggregated replication, active-control trials with documented dosage and validated outcome measures, and studies of how theoretical labels shape teacher decision-making.