Robert D. Tennyson

Offers a reflective overview of learning philosophies and theories underpinning instructional design and describes a “linking” approach that ties learning theory directly to research and practice. Emphasizes the need for designers to articulate clear theoretical foundations when creating effective learning environments.

Defines three core knowledge domains for effective instructional technologists—educational foundations, instructional systems development (ISD) methodology, and instructional development (ID) process—and argues that competence across these domains improves learning outcomes and reduces risk in technology‑based learning. Provides a worksheet for self‑assessment of competency.

Reviews MAIS as a design model linking knowledge acquisition (declarative, procedural, contextual) to specific instructional prescriptions (expository, practice, problem‑oriented) and linking knowledge employment (recall, problem solving, creativity) to complex‑dynamic and self‑directed strategies—each composed of empirically tested variables and conditions.

Presents a design theory for concept teaching grounded in a sustained program of empirical work. The model treats concept learning as two phases—forming conceptual knowledge and developing procedural skill—and organizes content‑structure and instructional‑design variables to select one of four strategy families. The article reviews supporting studies and explains how the strategies map to cognitive processes involved in concept acquisition.

Builds on two‑phase concept learning by testing response‑sensitive strategies that adapt the presentation form (expository vs. interrogatory) and the sequence of examples (generalization vs. discrimination rules). The adaptive‑selection plus generalization rule produced the strongest posttest and retention outcomes; fixed selection with generalization was weakest, highlighting benefits of adaptivity to learner error patterns.

Examines two design variables—response‑sensitive display‑time control and adaptive content sequencing—for coordinate concepts in CBI. Adaptive timing reduced the number of examples and total instructional time while improving posttest and retention performance compared with learner control. A combined generalization/discrimination strategy yielded superior learning efficiency versus either alone, consistent with a two‑phase concept‑learning theory.

Introduces MAIS as an integrated, adaptive CBI model with macro‑level expert‑tutor and learner models and micro‑level instructional variables that adjust in response to learner progress. Details how macro–micro interactions personalize instruction to improve learning efficiency and outcomes.

Tests whether manipulating example display time to match processing needs improves concept learning. With ninth‑grade students learning biological concepts via CBI, decreasing display time after incorrect responses (increasing information) and increasing time after correct responses (supporting processing) outperformed alternatives such as fixed timing or full learner control on both posttest and retention measures.

Describes MAIS, an AI‑enabled design architecture that iteratively adapts instruction to each learner’s moment‑to‑moment needs. Reviews the six empirically grounded design variables in MAIS and positions the system as a management‑systems approach to intelligent computer‑based instruction supported by a substantial research program.

With third‑grade learners on a mathematics concept, presenting best examples alongside the definition facilitated prototype formation, while combining expository and interrogatory examples best supported classification‑skill development and retention. Protocol data clarified which abstractions and descriptors students stored and used.

Evaluates whether combining expository statements of best examples (to support prototype formation) with interrogatory examples (to develop classification skills) enhances learning. Fourth‑grade learners receiving the expository‑interrogatory combination showed superior posttest and retention performance across concept‑attainment levels, supporting the two‑phase view of concept learning.

Synthesizes research directly related to concept teaching and connects it to experimental psychology literature. Proposes a four‑step design process: analyze content taxonomy; define concepts via critical attributes and select examples; arrange examples into rational sets through attribute manipulation; and order sets by divergence and difficulty. Identifies research gaps and directions for future work.

Investigates how manipulating positive and negative examples—and their similarity on irrelevant attributes—affects concept acquisition of adverbs among seventh‑graders. Results showed predicted patterns of correct classification, overgeneralization, undergeneralization, and misconceptions, and demonstrated that including negative instances is integral to effective concept instruction.

With undergraduates learning an infinite concept class in poetry, manipulations of exemplar/non‑exemplar probability, exemplar–non‑exemplar similarity, and exemplar divergence significantly affected correct classification and systematic error types (over‑ and under‑generalization, misconception), demonstrating how instructional instance design shapes concept learning outcomes.