Education · Latest edition · Issue 08 ·
Student Success
Two newly archived sources cover a UK staff–student perceptions study and U.S. advising controls. Neither demonstrates learning, retention or savings. The September 10 paper analyzes older survey data; Virginia Tech guidance is explicit February backfill. No cross-source patterns are asserted. Trial access failures and unresolved reporting inconsistencies limit outcome coverage; durable learning, disability-specific effects, advising impact and total cost remain gaps.
Read the edition Previous: Issue 07, September 12All Student Success editions
What this stream covers
Higher education teaching, advising, retention, accessibility, learning and student support. Evaluate durable learning and equity, not engagement alone. School-age learning belongs primarily to K12.
- Evidence records
- 2
- Cross-source patterns
- 0
- Also published September 13
- Campus OperationsCollege AthleticsEmergency ServicesK–12Local GovernmentNVIDIAPublic SafetyResearchState Government
- Subscribe
- Atom feed for Student Success
The latest edition did not claim a cross-source pattern; each of its records stands on its own.
Research through your lens
Every resource includes source evidence and takeaways for all three roles.
Evidence in this micro-vertical
28 resources
Follow the outcomes
28 resources across outcomes in your selection. Counts include all outcomes.
Refine by evidence type and topic
- Source
- Responsible Use of AI in Advising
- Published
- Living guidance last updated February 2026; exact day unknown
Virginia Tech ties advising AI use to approved accounts, consent and output review
The guidance requires approved institutional accounts, participant consent for interactions involving others, and advisor validation of generated work.
Limitations & uncertainty
February guidance is backfill, not a new September rollout. Institutional control statements are not independent assurance or a finding of legal compliance.
- Source
- Let’s Chat: Leveraging Chatbot Outreach for Improved Course Performance
- Published
- June 2026 manuscript version; exact day unknown
Course chatbot trial finds bounded grade gains with weaker adjusted evidence
Non-generative course outreach improved the A/B grade threshold, but evidence weakens after multiple-comparison correction and does not establish transferable learning.
Limitations & uncertainty
One institution and texting opt-ins; pooled numeric-grade effect nonsignificant. No strong spillover or subsequent-term effect; no formal cost-effectiveness analysis. The text's significance description needs qualification against Table 2. Grades are not a direct durable-learning test.
- Source
- Generative AI in Higher Education Teaching & Learning: Policy Framework
- Published
- December 2025, version 1.0; exact day unknown
Irish framework connects learning oversight, equitable access and data control
Guidance links teaching use to human accountability, equitable access, student data rights and ongoing review; it reports no measured implementation benefits.
Limitations & uncertainty
Normative Irish framework, not audit or outcome research. Controls require resources and enforcement; possession of a policy cannot establish compliance, equity or educational effectiveness.
- Source
- Student digital experience insights survey 2024/25: UK higher education survey findings
- Published
- September 2025; exact day unknown
UK student survey identifies access and AI-support gaps, with important sampling limits
Student-reported AI use sits alongside gaps in training and basic digital access; the survey does not measure AI learning effectiveness.
Limitations & uncertainty
Self-report, not learning or causal evaluation. One institution supplied 5,550 responses. The support-tools accessibility category combines AI with other tools, so it cannot establish AI-specific disability benefit.
- Source
- Still emerging: understanding Generative AI use in Higher Education
- Published
- September 10, 2026; survey conducted April–June 2024
UK survey exposes differing AI expectations without measuring learning
Staff and students differed in their expectations and perceptions of AI use; the study does not measure learning gains.
Limitations & uncertainty
Self-selection, low response, one school and 2024 data constrain current generalization. The underlying dataset is not openly released. Scale consistency cannot establish educational efficacy.
- Source
- UCF FinTech-AI Lab Is Shaping AI Talent
- Published
- September 9, 2026
UCF describes selective AI capstones with industry mentors; learning gains remain unmeasured
UCF describes an industry-mentored alternative to the fintech capstone. Educational and career benefits are operator claims without comparative outcomes.
Limitations & uncertainty
Selective promotional case, not causal evaluation. Productization is a possibility, not demonstrated deployment. Event date unknown.
- Source
- Guarded adoption of generative AI in higher education: high-achieving students, successful-student identity, and epistemic agency in a single-university mixed-methods survey
- Published
- September 9, 2026
Australian survey identifies selective AI use without establishing learning effects
Higher self-reported GPA was associated with less enthusiastic AI engagement; qualitative accounts describe selective, verification-intensive use. This is not evidence that avoiding AI improves grades.
Limitations & uncertainty
Self-reported achievement, self-selection, unknown response rate and one site. Composite reliability was weak; item-level checks help but do not establish causality. Successful-student identity is an interpretive label, not a validated construct. A 2025 snapshot, not current adoption prevalence.
- Source
- AI Tutoring Enhances Student Learning Without Crowding Out Reading Effort
- Published
- December 2025; exact publication day unknown
Tutor trial supports short-term learning, but delayed-access comparison is uncertain
An individually randomized experiment reports a 0.227 SD gain with AI access versus textbook-only study (SE 0.106). Immediate access exceeded delayed access by about 0.21 SD, but p=.066 qualifies the abstract's stronger significance language.
Limitations & uncertainty
Working paper, platform collaboration, short incentivized laboratory task; no delayed retention. Subgroup analyses are low-powered and unadjusted for multiplicity. Time to first prompt is an incomplete reading-effort measure.
- Source
- Guidance Over Adoption: Experimental Evidence on AI-Assisted Learning
- Published
- March 2026; exact publication day unknown
Chile trial separates adoption from learning: tutor-use guidance improves final-exam performance
Randomized encouragement increased tool adoption without detectable midterm improvement; separate tutor-use guidance improved final-exam outcomes. Table 3 reports a 0.218 SD intention-to-treat grade gain.
Limitations & uncertainty
Working paper; one course, partial participation, self-reported usage and peer spillovers. No delayed learning measure. Abstract rounds differently from Table 3; use the table estimate, not a stronger universal claim.
- Source
- Impacts of asynchronous learning modules on genAI competency in college students
- Published
- April 2026
Randomized university study finds 90-minute AI literacy modules improve some competencies but not critical analysis
A randomized study assigned 1,368 undergraduate and graduate students in 53 courses taught by 46 instructors to either no intervention or four self-paced modules totaling about 90 minutes. The modules significantly improved knowledge of how LLMs work, prompting skill, and self-efficacy beyond the control group, but did not significantly improve responsible-use knowledge or overall skill at analyzing AI output.
Limitations & uncertainty
The study occurred at one selective university with instructors who volunteered their courses, measured outcomes four days after access, and does not establish durable behavior change or safer real-world AI use. The output-analysis measure used a 174-student subset, and the intervention produced no detected gain in responsible-use knowledge or overall output analysis.
- Source
- We have had enough: thousands of University of Sydney staff walk off the job over AI and job security
- Published
- September 2, 2026
AI safeguards become a bargaining issue as roughly 2,000 university staff strike
About 2,000 University of Sydney staff joined a 24-hour strike amid enterprise bargaining disputes involving AI protections, workload fairness, and job security. The union sought enforceable safeguards in the employment agreement; the university said it supported many objectives but preferred to govern AI through institutional policies and maintained that the strike was premature.
Limitations & uncertainty
The report covers an active labor dispute, not an adjudicated finding of unsafe AI use. AI was one of multiple bargaining and trust issues, attendance estimates were reported rather than independently audited, and the internal trust result came from one faculty and a broadly worded statement. No AI system performance or educational outcome was evaluated.
- Source
- The Impact of Generative Artificial Intelligence on Student Creativity in Art and Design Education: A Meta-Analysis
- Published
- September 1, 2026
Meta-analysis finds a moderate positive effect on creativity in art and design education
A meta-analysis of 31 independent experimental, quasi-experimental, and correlational studies published from 2020 through 2026 found a moderate, statistically significant association between GenAI use and student creativity in art and design education, with an overall standardized effect of d = 0.61.
Limitations & uncertainty
The synthesis is limited to art and design education and combines experimental, quasi-experimental, and correlational evidence. The public abstract does not expose all heterogeneity and study-quality statistics, effects were smaller for K-12 students than graduate students, and a moderate average effect does not predict results for a particular curriculum or tool.
September review offers an integration map with substantial evidence-quality limits
The review maps AI use cases and integration depth; it does not estimate a pooled learning effect or establish that deeper technology integration is better.
Limitations & uncertainty
English-only, excludes grey literature, heterogeneous designs and an older search cutoff. Abstract's single Redefinition claim conflicts with multiple Table 1 labels; exact category counts are not reused.
- Source
- AI chatbot helps teach online-only psychology classes at Macquarie University
- Published
- September 1, 2026
University teaching chatbot scales rapidly while exposing unresolved learning and workforce tradeoffs
Macquarie's educator-configured Virtual Peer became part of weekly learning in two mandatory psychology units offered online. The AI activities were optional and used professor-supplied, checked material, while paid tutors still offered optional feedback sessions. The online format no longer included the prior optional weekly Zoom tutorials, prompting some students and staff to question whether AI was supplementing or displacing human teaching.
Limitations & uncertainty
The source is independent reporting rather than a formal evaluation. Usage and satisfaction figures are university-reported, student concerns are illustrative rather than representative, the activities were optional, and the reporting does not establish that AI caused staffing or modality decisions or changed learning outcomes.
- Source
- Ethically integrated generative AI for reading and vocabulary development in higher education: an experimental study of efficacy and learner perceptions
- Published
- September 1, 2026
Language-learning study reports reading gains, but later vocabulary advantage disappears and reporting is inconsistent
A 15-week, intact-class comparison reports reading benefits from guided multi-tool AI instruction. Table 4 shows no significant later vocabulary difference (p=.146). Broad efficacy language needs qualification.
Limitations & uncertainty
Nonrandom assignment, single setting, no delayed post-test, unisolated tool effects and inconsistent participant/statistical reporting limit confidence. The study describes January–May 2025 activity; no single event date is assigned.
- Source
- Epistemic dependence in AI-mediated learning
- Published
- August 29, 2026
Critical review offers questions for preserving learner judgment, not a validated dependency scale
The review distinguishes useful assistance from delegation that displaces learner judgment; it does not establish that frequent AI use causes harm.
Limitations & uncertainty
Nonexhaustive conceptual review; hypotheses from adjacent domains are not demonstrated educational effects. Detailed table endpoint failed, so no table-only claims are used; main-text methods and limitations were accessible.
- Source
- Layer-sensitive cognitive offloading in generative AI-assisted writing: supported performance and independent no-AI outcomes
- Published
- August 28, 2026
Writing study favors bounded support on independent tasks, with fragile class-level inference
Open collaboration produced the highest supported-writing mean; bounded support plus reflection led on independent Week 8 outcomes. The six-class design supports associations, not a definitive causal claim.
Limitations & uncertainty
Nonrandom intact classes; bundled reflection and delegation limits; same-course immediate near transfer only. Possible rater unblinding and demand effects. Raw data were not independently audited.
- Source
- AI in Texas: DIR Implementation of Laws from the 89th Legislature
- Published
- August 14, 2026
Texas turns AI legislation into shared governance and enablement services
Texas DIR reports implementing a legislative AI framework through a dedicated AI Division, government AI inventories, a code of ethics and heightened-scrutiny rules, a public-sector sandbox, model policy, certified awareness training, literacy programs, evaluation support, and cooperative contracts.
Limitations & uncertainty
DIR's update is self-reported government implementation evidence. Participation counts do not demonstrate safer systems, improved services, workforce productivity, or public value, and the long-term effect of the framework remains unmeasured.
- Source
- The regulation paradox: agentic AI, bounded autonomy, and self-regulated learning in higher education
- Published
- August 12, 2026
Perspective proposes explicit limits on educational agents' decision authority
The authors propose restricting agent initiative and returning planning and evaluation responsibility to learners; the proposal is not empirically validated.
Limitations & uncertainty
Conceptual mechanisms and proposed safeguards require testing. Cited studies span different educational contexts; their findings cannot be treated as direct evaluations of this framework.
- Source
- Generative AI Availability, Grades, and Student Satisfaction at a Large University
- Published
- July 23, 2026
U.S. university preprint finds no significant average grade effect, with important causal limitations
A university-scale observational analysis finds no average grade effect significant at 5% after accounting for pandemic disruption. This challenges universal grade-inflation claims without proving learning is unharmed.
Limitations & uncertainty
Preprint; grades are not direct learning measures. Grade parallel trends fail even before COVID, precluding strict causal interpretation. Exposure is inferred from syllabi; model error, grading changes and survey selection remain. Event period spans years.
- Source
- Using AI-based Learning Assistants in Higher Education: A Large-Scale Descriptive Analysis
- Published
- July 9, 2026
Large learning-assistant usage study highlights access confounding and inconsistent denominators
Observed adoption differs across groups, but course availability can confound comparisons; use logs do not establish learning benefit.
Limitations & uncertainty
Single month, single operator-affiliated study, inconsistent sample reporting, small subgroups and uneven course coverage. No causal comparator, interaction-quality evaluation or direct learning outcome.
Panel study links learning-centred AI use to academic functioning, not causal achievement gains
Learning-centred AI use predicted self-regulation and self-efficacy, with indirect associations to later engagement and procrastination. This is not evidence of improved grades or institutional retention.
Limitations & uncertainty
Observational, selective attrition, brief context-adapted scales and short intervals; weighting cannot remove unobserved selection. No repeated objective achievement measure.
- Source
- The Missing Evaluation Axis: What 10,000 Student Submissions Reveal About AI Tutor Effectiveness
- Published
- May 7, 2026
Programming tutor comparison finds stronger feedback uptake but mixed correct application
A misconception-focused tutor shows more feedback uptake, but correct application varies by assignment. This is not a learning-gain estimate.
Limitations & uncertainty
Nonrandom cross-semester comparison; cohort confounding, indirect attribution of edits and no delayed learning test. LLM judging has limited human validation. Optional helpfulness ratings cover about 38% of sampled submissions.
- Source
- AI chatbots in higher education: Comparing expectations to evidence
- Published
- April 17, 2026
U.S. course-chatbot trial finds no significant measured benefit; design limits matter
A course-grounded chatbot produced no significant treatment effects on interest, self-efficacy, eBook engagement or test achievement.
Limitations & uncertainty
One instructor/course; participation incentives shifted between tools; study-habit substitution was unmeasured. Table 2 uses doubled group counts; regression degrees of freedom require clarification before replication. No causal evidence that adding memory would improve learning.
- Source
- Scientific Reports
- Published
- February 6, 2026; version of record dated February 24
Engineering assistant study separates convenient help from demonstrated learning
Students valued convenient task support but expressed policy uncertainty; measured engagement does not establish learning gains.
Limitations & uncertainty
Abstract reports 71 participants, conflicting with methods' 65. Voluntary participation, one institution, self-report and novelty limit inference; external AI use is unobserved.
- Source
- Generative AI: product safety standards
- Published
- January 19, 2026
Education safety standards turn broad AI principles into product requirements
The Department for Education published a supplier-oriented baseline covering stated purpose, learner population, evidence claims, safeguarding, access control, testing, patching, privacy, and equality duties for generative AI products.
Limitations & uncertainty
This is normative guidance, not an evaluation of products or evidence that suppliers currently meet the requirements; several assurances depend on upstream providers and buyer verification.
- Source
- Streamlining Advising with Zoom AI Companion
- Published
- January 14, 2026
Utah describes consent and advisor review for AI appointment notes; benefits remain unmeasured
Utah describes explicit verbal student consent, the ability to stop AI Companion, and advisor correction before saving summaries in Navigate. Efficiency and record-quality benefits are objectives, not measured results.
Limitations & uncertainty
Operator account, not evaluation or full standards text. The November 21 approval mention omits the year, so eventDate is null. Training plans do not establish completed rollout.
- Source
- Short-Term Gains, Long-Term Gaps: The Impact of GenAI and Search Technologies on Retention
- Published
- July 10, 2025
Task-specific AI learning gains weaken at follow-up; reporting limits qualify the retention claim
ChatGPT improved immediate lower-order task assessment relative to control, but the study does not establish a general durable-learning advantage.
Limitations & uncertainty
Single site, post-assignment exclusions, fixed task order and restricted tools limit generalization. Cluster counts total 153 despite 152 volunteers; Task 2 prose conflicts with its table. Do not infer higher-order harm from nonsignificance. Model version and event dates are unspecified.
Stream editions
Each edition carries its own synthesis and evidence ledger.
September 13, 20261 edition
September 12, 20261 edition
September 11, 20261 edition
September 10, 20261 edition
September 9, 20261 edition
September 8, 20261 edition
September 7, 20261 edition
September 6, 20261 edition