Relied-on contributions and their limits
This edition connects formal and pragmatic explanation, empirical work on learning and interaction, and constructive practice through their contributions to a recipient’s task. The sources below support the stated distinctions and Methods. Worked examples in this framework are constructions unless explicitly reported as source observations.
| Source and access | Contribution adopted or adapted here | Limit on reliance |
|---|---|---|
| Galbraith and Khalifa, 2026, especially §§2, 5 | Interrogative treatment of understanding informs question selection and revision in EXD.1/.4. | Their theoretical focus does not settle every epistemic correctness or grasping question, nor certify an instructional intervention. |
| Fisher, 2022 | Natural consultations inform the joint organization and changing explanatory contributions in EXD.4. | Eleven consultations support the reported analysis; their phases are not a compulsory order or universal effect. |
| Axelsson, Buschmeier and Skantze, 2022, §2 | Contextual feedback distinctions support interpretation of contact, perception, understanding and attitude. | A surface signal alone does not identify a recipient’s internal state. |
| Social Explainable AI, 2026, ch. 5, ch. 12, ch. 13 | Explaining practices (§5.3), incremental explanation (§12.3.1–3) and adaptation (§13.2–3) inform co-construction. | Their AI-oriented realization is adapted by actual role and subject; interaction is not evidence of general learning benefit. |
| Liao and colleagues, 2021; Schmude and colleagues, 2025 | Elicit questions and purposes before choosing explanatory support; use a question bank for relevant clarification. | A design method or the latter study’s 24 participants and two contexts do not establish a complete universal question inventory. |
| Fichtel and colleagues, 2025 | Compare co-constructive interaction and objective understanding separately. | The study used two prompting conditions with Llama 3.1 70B, three topics and 277 retained participants. No significant mean objective-understanding advantage was found; this does not establish impossibility of a benefit in other conditions. |
| Relational-mapping study, 2023 | Principle–example and example–example mapping, with free and component-guided contributions, inform EXD.2. | Probability-learning conditions do not select one universally superior sequence or require unassisted failure. |
| Fiorella, 2023 | Explaining, visualizing and enacting make different contributions; modeled, focused and partial guidance inform EXD.2/.5. | Requesting an explanation from every recipient is not implied. |
| Mayer, 2024; Cromley and Chen, 2025 | Proximity, signaling, segmenting and relevant preparation inform EXD.3’s representation choices. | The synthesis covers 181 studies in 92 Mayer articles from 1990–2022. Its scope and moderation do not support a universal format winner or ban on repetition. |
| Zhang and Fiorella, 2024 | Error-focused explanation, correction and retry inform EXD.5. | The second experiment concerns 73 learners who made errors; correction and near-transfer findings do not establish far transfer. |
| Tetzlaff and colleagues, 2025; biochemistry study, 2025 | Task-specific prior knowledge and assistance conditions qualify instructional sequencing. | Sixty experiments in the synthesis do not prescribe automatic fading. The biochemistry study’s population-dependent near-transfer results do not equate course experience with the relevant schema or establish far transfer. |
| Mar and colleagues, 2021; Mengelkamp and colleagues, 2025 | Narrative/expository contrasts and comprehension–metacomprehension distinctions qualify expression and comparison choices. | Genre, content control, reading goal and measurement conditions constrain use. Memorability alone neither warrants the subject account nor chooses its form. |
Older books by Brown and Wragg (2001), LeFever (2012), Khalifa (2017) and McCain (2022) identify lines of instructional practice, explanatory craft and philosophical inquiry in the field’s background. Their publisher descriptions and contents supplied lineage orientation. Unread chapters are not premises for the operative Methods above; the later sources supply the named receiving contributions and limits.