Where it is used
Each entry identifies the volume, edition, chapter or appendix in which the work appears.
The evidence behind the series
Follow the research.
Explore every source.
Browse the scientific research and professional literature used across all seven volumes of The Science of Learning for Educators.
7book volumes
8author ranges
1searchable library
Search the collection
Find a paper, author or idea.
Choose an author range, search a title or concept, or narrow the collection by source type. Open any result to see where it appears in the books, read its evidence summary and follow available source links.
2 references
-
- Vol 5 RE: Chapter 15
generative artificial intelligenceai assessment scaleassessment validityhuman learning evidenceai-assisted assessmentassessment redesignacademic integrityPreprint Evidence summaryAt a glance
Perkins, Roe, and Furze revise the AI Assessment Scale as both a communication tool and a framework for redesigning assessment. It retains five levels, ranging from no generative-AI use through increasingly integrated use to AI exploration, but replaces the original traffic-light visual with neutral presentation so that levels are not mistaken for a hierarchy of quality.
Key ideas
Each level clarifies what students may do with AI and what evidence of human learning the task should preserve. The revision grounds the framework in social constructivism and assessment validity, incorporates feedback from implementation across school and higher-education settings, and addresses expanding multimodal capabilities. Practical vignettes illustrate how institutions and teachers can adapt the scale rather than impose identical rules on every subject or outcome. The authors warn against using it as a surveillance device, a universal policy, or a substitute for assessment design.
Why it matters
For educators, its strongest use is to make expectations explicit and reconsider whether a task still elicits valid evidence of the intended learning when AI assistance is available.
-
- Vol 4 RE: Chapter 26
generative artificial intelligenceself-directed learningpersonalized learning supportai literacyeducator guidanceacademic integrityinformation evaluationlearning technology researchPreprint Evidence summaryAt a glance
Roe and Perkins map research at the intersection of generative artificial intelligence and self-directed learning through a PRISMA-ScR-informed review of 18 studies published from 2020 to 2024. Their synthesis identifies four recurring themes: possible enhancement of self-directed learning, the continuing importance of educators as guides, personalization of learning support, and the need for caution.
Key ideas
Generative tools can provide rapid, on-demand explanations, feedback, and assistance adapted to a learner’s questions, potentially widening opportunities to plan and pursue learning independently. The evidence also raises concerns about accuracy, overreliance, academic integrity, privacy, and learners’ ability to judge generated information. The authors therefore reject the idea that AI makes teachers unnecessary; educators remain central in developing AI literacy, structuring tasks, and helping learners evaluate outputs. The review emphasizes that the evidence base is young, dominated by text-based systems and short-term studies.
Why it matters
It calls for longitudinal, empirical work on learning outcomes, diverse learner groups, and emerging multimodal tools before strong causal or general claims are made.
A transparent research trail
More than a bibliography.
What it contributes
Evidence summaries explain the central idea and why the source matters to educators.
How to access it
Entries link to lawful full text where available, or to a publisher, library or purchase page.
Edition abbreviations
- TE
- Teacher Edition
- PE
- Practice Edition
- RE
- Research Edition
Citation checks
References are checked against publisher records, scholarly indexes, DOI or ISBN metadata, repositories and author records where available. An entry marked [Citation not verified] preserves the wording found in the relevant volume without attributing findings to an unconfirmed work.
From sources to a coherent pathway