The Science of Learning Weekly Dispatch #1

Your weekly round-up of the science of learning news and research

This is the first edition of a weekly brief on the latest research, insights, and developments in the science of learning. Each week, The Learning Dispatch will bring you a concise roundup of newly published studies, key debates in education, and practical applications of cognitive science. The aim here is to keep you informed and help bridge the gap between research and classroom practice without the fluff.

This Week in the Science of Learning

• Students’, Teachers’, and Parents’ Knowledge About and Perceptions of Learning Strategies Apparently we all know effective learning strategies (like retrieval practice & spaced study) but overrate familiar methods like highlighting.

  Key finding: Simply knowing what works isn't enough, we need concrete examples of how to implement strategies effectively. As ever, time & effort remain barriers to using best practices.

• Does Instruction-First or Problem-Solving-First Depend on Learners’ Prior Knowledge? This one is surprising - a new article (in press) challenges aspects of cognitive load theory by examining the effectiveness of different instructional sequences (explicit instruction before problem-solving vs. problem-solving before instruction). Here's the really odd thing: it suggests that problem-solving first benefits novices, while explicit instruction first benefits more advanced students.

• People who have more science education rely less on misinformation—Even if they do not necessarily follow the health recommendations people with higher science education and motivation (personal relevance) are less likely to rely on misinformation but do not necessarily follow expert recommendations. Science education improves information evaluation and argument complexity, but not adherence to recommendations, suggesting that informed reasoning is distinct from behavior.

• The Testing Effect in the Lecture Hall: Does it Transfer to Content Studied but Not Practiced? practice testing improves long-term retention but only for explicitly tested content—there was no transfer effect to untested material from the same lecture. To maximize learning, educators should ensure that key content is directly included in quizzes rather than assuming broader benefits from retrieval practice alone.

• The role of feedback and working memory for goal-related monitoring and goal revision Feedback improves children's goal monitoring and revision, especially for those with weaker working memory, helping them set realistic goals and adjust their performance expectations. Without feedback, children tended to overestimate their performance and set overly ambitious goals, potentially undermining motivation.

• What is creative in childhood writing? This study examined how linguistic features in children's stories predict human-rated creativity. Children’s creative writing is judged more highly when stories are longer, include diverse and original words, and connect divergent ideas, while logical writing depends more on grammatical accuracy, syntactic simplicity, and coherence. Automated linguistic measures can aid creativity assessment but cannot fully replace human judgment.

• A scoping review of research on individual differences in the testing effect paradigm retrieval practice benefits learning regardless of individual differences like working memory or motivation. Individual differences such as working memory, motivation, and test anxiety do not significantly alter its effectiveness, suggesting that retrieval benefits are broadly applicable

• Chatbots in education: Hype or help? A meta-analysis chatbots can help students learn, especially in STEM subjects and when used over a longer period. Text-based chatbots work better than speech-based ones, but the overall impact is smaller than some studies suggest due to publication bias. The key message? Chatbots can be useful, but they’re not a magic fix and how they’re integrated into teaching matters more than just having them.

• When struggling readers meet the screen – A secondary analysis of ePIRLS 2016 data Struggling readers tend to do worse when they use digital devices for schoolwork. In contrast, for students who are better at print reading, using digital devices was linked to better digital reading comprehension.

• ChatGPT Goes to College: Exploring Student Perspectives on Artificial Intelligence in the Classroom College students primarily use ChatGPT for information gathering due to its convenience rather than for entertainment, and they generally recognize unethical uses as academic misconduct, highlighting the need for clear institutional AI policies in higher education.

Science of Learning in the news roundup

Louisiana’s NAEP Score — A Victory for the Science of Reading?

There Really Was a ‘Mississippi Miracle’ in Reading. States Should Learn From It

Spellcheck Won’t Cut It. Here’s Why Kids Need Spelling Instruction

The buzz around teaching facts to boost reading is bigger than the evidence for it (a contentious article. See Robert Pondiscio’s response here.)

Watch: Are China’s International Test Scores Too Good to Be True?