The exact text we saved

This is the version exactly as it was saved, character for character. The SHA-256 hash below is a 64-character code worked out from that text. Run the same text through SHA-256 and you get the same code; change a single letter and the code comes out completely different. Each version also stores the previous version’s hash, which chains them together so no version can be swapped out quietly.

TRUVACE RECORD VERSION
record: TRV-2026-2811
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-11-12T17:27:15.029332Z
status: published
lens: g_space
sector: education
headline: Exploring the Impact of Generative AI-Powered Conversational Agents on Student Learning: A Systematic Review and Meta-Analysis Grounded in Activity Theory
dek: Grounded in activity theory, this study conducted a systematic review and meta-analysis, following PRISMA guidelines, to evaluate the impact of generative AI-powered conversational agents on students’ learning outcomes in both cognitive and non-cognitive skills. Twenty-seven empirical studies published between 2022 and 2025 were analyzed. The results revealed that generative AI-powered conversational agents had moderately positive effects on both cognitive ( g = 0.462, p < .001) and non-cognitive skills ( g = 0.…
gain_title: A systematic review and meta-analysis of 27 studies published from 2022 to 2025 shows students who learned with generative AI chatbots gained moderately in cognitive skills (g = 0.462) and non-cognitive skills (g = 0.519).
problem_title: (none)
trace_subject: (none)
gain_reading: A systematic review and meta-analysis of 27 studies published from 2022 to 2025 shows students who learned with generative AI chatbots gained moderately in cognitive skills (g = 0.462) and non-cognitive skills (g = 0.519).
gain_evidence: moderately positive effects on both cognitive ( g = 0.462, p < .001) and non-cognitive skills ( g = 0.519, p < .001)
problem_reading: (none)
problem_evidence: (none)
quick_read: A systematic review and meta-analysis following PRISMA guidelines pooled 27 empirical studies published between 2022 and 2025. It examined how generative AI-powered conversational agents affect students' learning outcomes.

The pooled effect was moderately positive for cognitive skills (g = 0.462, p < .001) and slightly larger for non-cognitive skills (g = 0.519, p < .001).

Moderator analyses found that educational level, knowledge type, intervention duration, agent role and instructional mode were significant or marginally significant factors, so results differ by classroom context.
limitation: 
tag: Evidence-backed gain
key_points: The review followed PRISMA guidelines and analysed 27 empirical studies published between 2022 and 2025. | The pooled effect on cognitive skills was g = 0.462 (p < .001). | The pooled effect on non-cognitive skills was g = 0.519 (p < .001). | Educational level, knowledge type, intervention duration, agent role and instructional mode were significant or marginally significant moderators. | The analysis is grounded in activity theory and covers both cognitive and non-cognitive learning outcomes.
rundown: The authors synthesised empirical studies from 2022 to 2025, a period covering the spread of generative AI chatbots in education. They assessed outcomes in two groups: cognitive skills and non-cognitive skills.

Both groups showed moderate positive pooled effects, with g = 0.462 for cognitive and g = 0.519 for non-cognitive skills, both at p < .001.

Moderator analyses indicated that effects depend on context, including educational level, type of knowledge, how long the intervention ran, the role the agent played and the mode of instruction.
sources:
- peer_reviewed | Journal of Educational Computing Research | https://doi.org/10.1177/07356331261433396 | 2026-03-20
prev: 0000000000000000000000000000000000000000000000000000000000000000

SHA-256 hash

94d9a6a90fbd75fd38a9ef0c90c9ec95eb3b42163f866506806543fc07bd28b5

Previous version’s hash

0000000000000000000000000000000000000000000000000000000000000000