{"resourceId":"german-secondary-srl-genai-rct-2026","versions":[{"version":"external-dc786f57d91cf230cd657dd0615b9319314155ccde94b0f921d5b791da6444c8","resource":{"id":"german-secondary-srl-genai-rct-2026","title":"Classroom trial finds limited benefits from specialized self-regulation prompts","organization":"Tim Fütterer and colleagues","sector":"K–12 secondary education","geography":"Baden-Württemberg, Germany","publishedAt":"April 14, 2026; classroom study April–May 2025","publicationDate":"2026-04-14","eventDate":null,"sourceName":"Enhancing School Students’ Self-Regulated Learning through Generative AI Support: A Randomized Controlled Trial","sourceLabel":"Peer-reviewed classroom randomized trial","sourceUrl":"https://link.springer.com/article/10.1007/s10648-026-10133-8","evidenceClass":"academic-research","outcomeClass":"mixed","topics":["knowledge-work","developers-agents","data-security","accessibility-workforce","operating-model"],"finding":"Specialized prompts produced no significant advantage over the AI control in knowledge, effort or strategy use. Utility-value trajectories favored motivational over strategy prompting, not over control.","sledRelevance":"Interpretation: Newly archived counterevidence for fall tutoring decisions; German secondary-school findings require local curriculum and learner validation.","evidence":"371 Grade 7–9 students; individual randomization; six 45-minute sessions including pre/posttests and four learning sessions. Mixed-effects analyses compared GPT-4o conditions. Posttest attrition was 35%.","architectureImplications":"Interpretation: version prompts, model and task content together; evaluate the complete learner workflow. Autonomous agents, local hosting and hybrid infrastructure were not compared.","governanceImplications":"Interpretation: define knowledge and motivation outcomes separately before the pilot; avoid selecting exploratory subgroups after results.","securityPrivacyImplications":"Interpretation: minimize learner context, restrict transcript access, document retention and prohibit unrelated reuse; local privacy review remains necessary.","caveats":"Convenience sample, immediate outcomes, attrition and measurement limitations. Control also used AI; no AI-versus-no-AI estimate. Supplementary data were not independently reanalyzed.","streamIds":["k12"],"roles":{"sales":"Interpretation: A curriculum director considering a study-support assistant needs evidence that students learn, beyond finding the interaction appealing. Ask which independent assessment will be used, whether the comparison includes ordinary instruction, and who supports disengaged learners. Offer a bounded evaluation design for one subject and grade. The value hypothesis is a clearer continuation decision. The trial's limited contrasts cannot support a general claim that AI improves self-regulation or that it never works. Treat the German setting as a reason to validate locally, and confirm purchasing needs directly rather than inferring an opportunity.","engineering":"Interpretation: For a classroom proof of value, isolate task materials, prompts and learner identity from unrelated student records. Establish a controlled release process and a teacher fallback before adding a chatbot to the learning platform. Prerequisites include approved curriculum, testable access rules and permission to evaluate the configured model. Use synthetic cases to test answer leakage and cross-student disclosure. Compare independent pre/post and delayed assessments with an ordinary-instruction baseline, while recording missing responses and review effort. These proposed tests address uncertainty; they do not reproduce the reported study or establish suitability for autonomous instruction.","delivery":"Interpretation: Assign the curriculum evaluation lead as owner, with classroom teachers, IT, accessibility and privacy staff responsible for defined checks. Reserve lesson time, prepare accessible alternatives and train teachers to identify frustration without taking over assessment responses. Approve the analysis and stopping rules before collecting data. Proposed acceptance: every participant has a permitted route through the activity, missing assessments are reported, and delayed performance and staff effort are reviewed before expansion. These are proposed criteria, not observed results. Risks include selective reporting, attrition and deploying a changed model under the original evidence."},"retrievedAt":"2026-09-08T03:01:24Z","enrichedAt":"2026-09-08T03:01:51Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Interpretation: test reading demands, accommodations and teacher intervention time; text interaction may impose uneven burdens.","procurementImplications":"Interpretation: require evidence for the actual configured service and exportable evaluations; no savings or retention guarantee is supported.","operatingModelImplications":"Interpretation: curriculum leads evaluation, teachers handle learner support, and IT controls versions and access.","updateExplanation":"Not an archive repeat: absent from all 20 K12 records and the full-library candidate search. Newly added context for fall 2026 decisions; not represented as new September 7 publication.","sourceVerification":{"openedUrl":"https://link.springer.com/article/10.1007/s10648-026-10133-8","referenceExcerpt":"No clear condition differences emerged for other learning outcomes.","promptVersion":"sled-research-v3.1","model":null,"basis":"agent-reported inspection"}}}]}