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-2026-2757
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-11-12T17:20:09.988955Z
status: published
lens: g_space
sector: education
headline: Problem-based and project-based learning as promising frameworks for generative AI-supported education: Emerging evidence from a systematic review and three-level meta-analysis
dek: As AI advances, specifying which competencies learners will need grows harder, making it urgent to identify stable approaches for cultivating competencies that remain valuable regardless of which AI tools emerge next. Problem-based learning (PBL) and project-based learning (PjBL) may provide one suitable learning context for this: by integrating generative AI (GenAI) into authentic problem solving and meaningful production, they may help cultivate the judgment to evaluate AI outputs and the capacity to coordinat…
gain_title: A meta-analysis of 22 controlled studies (66 effect sizes) found generative AI used in problem- and project-based learning raised learner outcomes by g = 0.819 overall, falling to g = 0.378 after adjusting for small-study effects.
problem_title: (none)
trace_subject: (none)
gain_reading: A meta-analysis of 22 controlled studies (66 effect sizes) found generative AI used in problem- and project-based learning raised learner outcomes by g = 0.819 overall, falling to g = 0.378 after adjusting for small-study effects.
gain_evidence: A three-level meta-analysis of 22 controlled studies (66 effect sizes, January 2023-April 2026) found significant effects across four learner-internal outcome domains | yielding an overall effect of g = 0.819
problem_reading: (none)
problem_evidence: (none)
quick_read: A three-level meta-analysis pooled 22 controlled studies, with 66 effect sizes published between January 2023 and April 2026, on generative AI inside problem-based and project-based learning. It found significant effects in four learner-internal domains and an overall effect of g = 0.819.

The headline figure is probably inflated. Sensitivity analyses showed small-study effects, and the PEESE correction cut the estimate to g = 0.378, still a positive but much smaller effect.

Product and solution performance, a traditional focus of these teaching methods, was modeled separately and showed a very large pooled estimate of g = 1.958. Its prediction interval crossed zero, so results across settings remain highly uncertain, and the authors call the evidence promising but preliminary.
limitation: Small-study effects suggest the unadjusted effect may be inflated, and the prediction interval for product and solution performance crossed zero, so the authors call the evidence promising but preliminary.
tag: Evidence-backed gain
key_points: The analysis covered 22 controlled studies and 66 effect sizes from January 2023 to April 2026. | Four outcome domains showed significant effects: learning achievement and skill performance, higher-order thinking, affective and motivational outcomes, and collaboration and communication. | The overall pooled effect was g = 0.819 before any bias adjustment. | After correcting for small-study effects with PEESE, the estimate dropped to g = 0.378. | Pooled effects differed significantly by participant number and by type of generative AI tool. | Peer collaboration as a moderator showed only a marginal trend. | Product and solution performance had a pooled estimate of g = 1.958, but its wide prediction interval crossed zero.
rundown: The authors describe the work as an early-stage synthesis of a nascent field. Their premise is that placing generative AI inside authentic problem solving and meaningful production may build the judgment to evaluate AI outputs and to coordinate AI use across complex workflows.

The study used a three-level meta-analysis to handle multiple effect sizes per study. It pooled four learner-internal outcome domains: achievement and skills, higher-order thinking, affective and motivational outcomes, and collaboration and communication.

Sensitivity checks pointed to small-study effects, meaning the unadjusted g = 0.819 may overstate the benefit. The PEESE specification gave g = 0.378. Exploratory moderator analyses found that estimates varied significantly with participant number and GenAI tool type.

Product and solution performance was analysed separately. Its pooled g of 1.958 may reflect AI-augmented task performance rather than learning, and the prediction interval crossed zero, indicating substantial uncertainty.
sources:
- peer_reviewed | Acta Psychologica | https://doi.org/10.1016/j.actpsy.2026.107734 | 2026-09-08
prev: 0000000000000000000000000000000000000000000000000000000000000000

SHA-256 hash

743b2dfdc626f2d61a66f5be32f1fd2bb54f34dbd7f166fdb9e90a8f9341191d

Previous version’s hash

0000000000000000000000000000000000000000000000000000000000000000