A higher proportion of untaught features is learned when teaching via sentence-selection compared to natural language¶
Claim · All claims
Evidence · 1 study · causalr1·q3
Currency · newest study 2025 · not yet checked against its sources · not reviewed by a person
Design limits · Rachel Love 2025: the intervention came bundled with another change (read from the study by two models, 2026-10-27)
Subclaims¶
Reading the evidence codes
q is the design tier, i the impact, n the sample, and the kind with its r
the rigour judged within that kind: what each code means.
This claim: no standardized effect is recorded in observations/ for it yet, so no benchmark reading is possible.
q3 i? Participants in the sentence-selection condition learned a significantly higher proportion of features they did not teach (M = 23.6%) compared to the natural language condition (M = 17.3%). → Rachel Love 2025
Evidence¶
Rachel Love 2025¶
Rachel Love, Edith Law, Philip R. Cohen, Dana Kulić. (2025). Teaching a Conversational Agent using Natural Language: Effect on Learning and Engagement. Int J Artif Intell Educ. https://doi.org/10.1007/s40593-025-00461-1
q3 · i? · causal · r1
Second exploratory two-way repeated measures ANOVA on untaught-feature learning. Sentence-selection produced significantly higher proportional learning of features never explicitly taught to the agent, suggesting incidental exposure benefits from reading full sentences.
"participants in the sentence-selection condition learned significantly more features they did not teach compared to the natural language condition (sentence-selection M = 23.6% (SD = 14.4); natural language M = 17.3% (SD = 10.3)), F(1, 39) = 7.51, p = .009."
Discussion¶
Related Claims¶
- A higher proportion of taught features is learned when teaching via natural language compared to sentence-selection — related
- Teaching via paraphrased natural language is perceived as more helpful for learning than sentence-selection, but this perception is not reflected in actual learning gains — related
- Teaching a teachable agent via paraphrased natural language produces no significantly different learning gains compared to teaching via sentence-selection over a fixed interaction — related
- Natural language teaching takes significantly longer per input and yields significantly fewer teaching attempts and fewer features taught than sentence-selection — related
Page metadata
| Field | Value |
|---|---|
type |
claim |
id |
sentence-selection-higher-proportion-untaught-learned |
title |
A higher proportion of untaught features is learned when teaching via sentence-selection compared to natural language |
description |
A higher proportion of untaught features is learned when teaching via sentence-selection compared to natural language |
status |
review |
evidence_strength |
moderate |
newest_evidence |
2025 |
generated |
by process:wiki-ingest on 2026-10-27 |
trust |
unverified — no one has confirmed this page's content |
sources |
1 — rachel-love-2025 |
Frontmatter is the page's machine-readable record — see the Schema & Guide.