🔬 Learning research, plainly by Ask Poppy

All summaries

Newest research first. Classics are marked as such.

SummaryYearStudy typeRead
With AI, people scored better on reasoning puzzles but overestimated how well they did2026Two online experimentsAbstract only
Unrestricted ChatGPT help raised practice scores but lowered exam scores2025Preregistered randomized controlled trial (field experiment)Full text read
Six weeks of teacher-guided AI tutoring improved English results in Nigeria2025Randomized controlled trialAbstract and parts of the full text read
A carefully designed AI tutor beat an active-learning physics class on a short-term test2025Randomized controlled trial with a crossover designFull text read
"Your Brain on ChatGPT": what the MIT essay-writing study did and didn't show2025Lab experiment with EEGFull text read
Supervised AI tutoring matched human tutors in UK maths classrooms2025Exploratory randomized controlled trialFull text read
ChatGPT made research feel easier but students' reasoning was weaker than with Google2024Randomized experimentAbstract only
An AI coach for human tutors raised topic mastery, most for weaker tutors2024Preregistered randomized controlled trialFull text read
Reading aloud helps you remember a text, but not understand it better2023Four lab experimentsAbstract only
Homeschooled teens in a large US cohort: more volunteering, fewer college degrees2021Longitudinal cohort study (observational)Full text read
Which kinds of motivation go with doing well: a meta-analysis of 223,209 students2021Meta-analysisAbstract only
Quizzing boosts classroom learning: a meta-analysis of 222 studies2021Systematic review and meta-analysisAbstract only
Prompting learners to explain to themselves helps: a meta-analysis of 64 reports2018Meta-analysisAbstract and parts of the full text read
Spacing beats cramming: the classic review of 317 experiments2006Review and meta-analysis (landmark classic)Full text read
The classic study: testing yourself beats rereading after a week2006Lab experiments (landmark classic)Abstract only