A.15.11:11 - SoTA-Echoing

The practice question is how to make a useful method available before a recipient asks for it, without imposing more burden than the contribution warrants. The selected answer is a methodological synthesis: arrange an encounter, expose the relation, preserve access and prerequisites, and compare use with non-intervention. No single cited experiment validates this complete method across domains.

Adapt relational comparison; reject a hint as proof of spontaneous use. Gick and Holyoak’s Schema Induction and Analogical Transfer (1983) remains a historical experimental anchor for this distinction. Their comparison of example structures informs §4.2; their prompted and unprompted conditions inform §4.5. A principle or diagram added to one story did not reliably produce the same benefit as their richer comparisons. These bounded tasks do not prescribe an example quota or establish transfer to all work. Relational comparison is worth its added effort when a simple pointer leaves applicability unclear.

Adapt cue placement and cost comparison. Tsai and Gilbert’s Strategic reminder setting for time-based intentions (2025) examines short laboratory intentions, reminder effort and cue visibility. It informs §§4.3 and 4.6: an external cue changes both the remembering task and the ongoing work. The serious alternative is using the support already available. These experiments concern known intentions and cannot establish discovery of an unnoticed method. Use their contribution after identifying a candidate occasion; do not generalize their brief delays or participant exclusions into everyday guarantees.

Adopt separate questions about immediate assistance and later use. Teo and colleagues’ AI Assistants Overassist (2026 preprint), §§3.1–3.5, separates intervention timing, immediate results and performance on a related problem under different retained contexts. It informs §4.5 and the comparison with no intervention in §4.6. A supplied answer may be the right immediate contribution, while development requires a different comparison. The simulated students and context manipulations do not establish parameter learning or human acquisition; the arrangement still needs examination in its receiving work.

Adapt a checklist to actual execution. Degani and Wiener, Human Factors of Flight-Deck Checklists: The Normal Checklist (1990), §§3.1.4 and 5.2.2, distinguishes doing items from verifying their state and describes a correct response being given without the corresponding verification. This historical aviation analysis informs §4.3 and §5.4: choose a real occasion, inspect the relevant condition, and give an unresolved item a continuation. A publication’s conformance list and a usable execution aid have different jobs. The aviation method does not make any finite list comprehensive for all work; an unlisted concern retains its own path into inquiry.

Reopen the choice when a simpler entry works just as well, a cue repeatedly misfires, changing work defeats its relevance, or a stronger subject method offers the same gain with less disruption. Human learning and automatic recommendation require their own more specific evidence when those are the intended claims.