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.
6 references
-
- Vol 5 RE: Chapter 12
learning analyticsgenerative artificial intelligencepersonalized interventionspedagogical decision makingaffective inferencestudent data privacyhuman oversightmachine-generated interpretationsPreprint Evidence summaryAt a glance
Sajja, Sermet, Cwiertny, and Demir describe a prototype learning-analytics tool that uses GPT-4 to help instructors interpret student questions and other learning signals.
Key ideas
The system classifies features such as topic, cognitive level using Bloom’s taxonomy, curiosity, confusion, stress, and preferred study approaches, then aggregates them into dashboards intended to show engagement and learning progression. The paper explains the architecture, learning-management-system integration, synthetic-data testing, and a faculty survey about perceived usefulness and risks. Respondents saw potential for timely instructional adjustment and personalized intervention but raised concerns about privacy, security, reliability, bias, and the validity of machine-generated interpretations. The study is exploratory: it does not demonstrate improved learning outcomes, and inference about affect from text requires careful validation.
Why it matters
Its contribution is a concrete design case showing how generative AI might augment pedagogical decision making while making governance, human oversight, transparent metrics, and protection of student data central requirements rather than afterthoughts.
-
- Vol 7 RE: Chapter 10, App D
learning analyticsstudent expectationsstudy planningself-assessmentadaptive recommendationspersonalized feedbackdata privacyJournal article Evidence summaryAt a glance
Schumacher and Ifenthaler investigate what university students want from learning-analytics systems and whether they believe particular features would support their learning. An exploratory qualitative study with 20 students generated a set of expected functions, followed by a quantitative study in which 216 students rated 15 features, their willingness to use them, and perceived learning value.
Key ideas
Students especially valued tools for planning and organizing study, self-assessment, adaptive recommendations, and personalized analysis of learning activity. These expectations map onto phases of self-regulated learning, including preparation, performance monitoring, and reflection. Participants saw potential benefits in timely feedback and individualized support, yet also worried that analytics might become invasive or reduce autonomy. The findings show why dashboard design should begin with learner needs and educational theory rather than the data that happen to be available.
Why it matters
The authors recommend aligning features with self-regulation, feedback, and instruction, while making data practices and purposes transparent. Meaningful user involvement is therefore both a design requirement and a condition for acceptance.
View or buy the cited work >>Read a similar article free online >>
-
- Vol 4 RE: Chapters 12, 18, App A, App C
self-regulation promptslearning analyticstrace datatransfer performanceadaptive promptingdigital learning environmentsinstructional contextJournal article Evidence summaryAt a glance
This quasi-experimental study examines whether prompts embedded in a digital learning environment can activate self-regulation strategies and whether learning-analytics trace data can explain performance. A sample of 110 higher-education students received prompts targeting cognitive, metacognitive, motivational, or resource-management processes, or participated in a control condition.
Key ideas
The researchers compared declarative knowledge, transfer, learner perceptions, and recorded online behavior. Prompting produced small effects on declarative knowledge and transfer, and prompted groups interacted with the environment differently from the control group. However, the available trace measures did not explain transfer performance sufficiently, illustrating the gap between logging observable clicks and inferring the cognitive or motivational processes that matter for learning. The authors conclude that prompts have potential as lightweight support for self-regulated learning, but learning analytics must be more closely tied to theory and instructional context.
Why it matters
They call for studies of adaptive rather than fixed prompts, richer interpretations of trace data, authentic learning settings, and closer attention to how individual learners perceive and respond to interventions.
View or buy the cited work >>Read a similar article free online >>
-
- Vol 4 RE: App C
self-regulated learningmetacognitive monitoringmotivational regulationaffective regulationhelp seekingcollaborative regulationlearning analyticsEdited book Evidence summaryAt a glance
This edited volume treats self-regulated learning as the active coordination of cognition, metacognition, motivation, affect, behavior, and environmental resources in pursuit of goals. Models differ in terminology but commonly describe recurring phases of task interpretation and planning, strategic performance and monitoring, and reflection that informs adaptation.
Key ideas
Contributions examine efficacy beliefs, goals, values, emotion, help seeking, collaboration, development, and differences across subject domains. They also show why self-report questionnaires alone cannot capture regulation as it unfolds; think-alouds, traces, diaries, microanalysis, and performance data reveal different parts of the process. Instruction can make strategies explicit, model conditional choices, provide feedback on progress, and gradually shift responsibility to learners. Technology may support prompts and visualization, but dashboards or prompts are not automatically regulatory.
Why it matters
The collection’s central message is that successful learners do not merely possess strategies: they decide when and why to use them, monitor consequences, and revise action within particular tasks and social contexts.
View or buy the cited work >>Read a similar article free online >>
-
- Vol 4 RE: Chapter 6
self-regulated learningsocially shared regulationcollaborative learningadaptive learning environmentslearning analyticsfeedback scaffoldinghelp seekinglearner agencyJournal article Evidence summaryAt a glance
Sharma, Nguyen, and Hong systematically review empirical studies connecting adaptive digital learning environments with self-regulated and socially shared regulation in collaborative learning. Their search and coding identify seven recurring objectives: feedback and scaffolding, regulatory skills and strategies, learning trajectories, collaborative processes, adaptation and regulation, self-assessment, and help seeking.
Key ideas
Adaptive systems can personalize prompts, representations, feedback, and paths, while collaboration introduces collective planning, monitoring, control, and reflection that cannot be inferred from individual traces alone. The literature uses heterogeneous theoretical models, settings, technologies, and measures, limiting cumulative conclusions. Important gaps include few informal-learning studies, weak theoretical convergence between individual and shared regulation, and difficulty monitoring how regulation is coordinated across group members. The authors call for clearer constructs, multimodal and process-sensitive measures, and designs that support rather than automate learner agency.
Why it matters
The review provides a map of opportunities but does not establish that adaptation by itself improves regulation or collaboration.
-
- Vol 5 RE: Chapters 10, 13, 15, App A
learning analyticseducational data miningpredictive modelinghuman sense-makingactionable analyticsdata-informed decision makingprivacylearner agencyConference paper Evidence summaryAt a glance
Siemens and Baker compare learning analytics and educational data mining, two overlapping communities using educational data to understand and improve learning. Educational data mining had tended to emphasise automated discovery, modelling, prediction, and adaptation at the level of learners and software, drawing strongly from computer science.
Key ideas
Learning analytics had placed relatively greater emphasis on human interpretation, sense-making, organisational decision making, social networks, and the needs of instructors and institutions. The authors argue that the differences are productive but that limited communication risks duplicated work and fragmented standards. They propose collaboration around shared research methods, tools, data, theory, conferences, and ethical challenges while preserving complementary perspectives. The short position paper does not present a new empirical evaluation; its contribution is field-building and agenda setting. For educational practice, analytics should connect technically credible models with actionable human judgement.
Why it matters
Predictions alone do not improve learning, and dashboards without theory or intervention can mislead. Researchers and institutions need transparent definitions, validation across contexts, attention to privacy and agency, and evaluation of whether data-informed actions actually benefit learners.
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