The research behind HeroMemo
HeroMemo is built on some of the best-tested findings about how children learn and remember. Here is the research, what each study really measured, and how we use it. Every paper links to its source.
Learning at the right level, about what they love
Teaching that adapts to each learner, like a good tutor, beats the same lesson for everyone. Linking lessons to a child's own interests helps too.
Students taught one-to-one, with checks that each step was mastered, scored about two standard deviations above students in a regular class: the average tutored student did better than 98% of the class.
What was measured: Small studies from the early 1980s. Later, larger reviews found smaller, but still large, gains.
Bloom (1984). The 2 Sigma Problem: The Search for Methods of Group Instruction as Effective as One-to-One Tutoring. Educational Researcher, 13(6), 4–16.
Read the paper (opens in a new tab)Human tutoring raised results by 0.79 standard deviations over no tutoring, and step-by-step computer tutors by 0.76: nearly as much.
What was measured: A review of controlled studies comparing kinds of tutoring, mostly in science and maths.
VanLehn (2011). The Relative Effectiveness of Human Tutoring, Intelligent Tutoring Systems, and Other Tutoring Systems. Educational Psychologist, 46(4), 197–221.
Read the paper (opens in a new tab)Across 50 controlled evaluations, intelligent tutoring systems raised test scores by a median of 0.66 standard deviations over regular teaching: from the 50th to the 75th percentile.
What was measured: Meta-analysis of 50 evaluations. Gains were larger on tests written for the course than on standardized tests.
Kulik & Fletcher (2016). Effectiveness of Intelligent Tutoring Systems: A Meta-Analytic Review. Review of Educational Research, 86(1), 42–78.
Read the paper (opens in a new tab)Mastery learning, where each learner moves on only after mastering a step and gets help when stuck, improved exam results. The gains were strongest for the weaker students in a class.
What was measured: Meta-analysis of 108 controlled evaluations in schools and colleges.
Kulik, Kulik & Bangert-Drowns (1990). Effectiveness of Mastery Learning Programs: A Meta-Analysis. Review of Educational Research, 60(2), 265–299.
Read the paper (opens in a new tab)Teaching that helps beginners can slow down more advanced learners, and the other way round. Instruction works best when it matches what each learner already knows.
What was measured: A review of experiments on how the best way to teach changes with experience.
Kalyuga, Ayres, Chandler & Sweller (2003). The Expertise Reversal Effect. Educational Psychologist, 38(1), 23–31.
Read the paper (opens in a new tab)Children who learned maths in a game personalized with their own name and favourite things, with some choices to make, were more motivated, more engaged, and learned more.
What was measured: An experiment with elementary school children using a maths computer game.
Cordova & Lepper (1996). Intrinsic motivation and the process of learning: Beneficial effects of contextualization, personalization, and choice. Journal of Educational Psychology, 88(4), 715–730.
Read the paper (opens in a new tab)Students whose algebra problems were rewritten around their own interests, such as sports, music or games, solved more problems correctly, especially harder ones, and learned more.
What was measured: Algebra students using an adaptive computer tutor.
Walkington (2013). Using adaptive learning technologies to personalize instruction to student interests: The impact of relevant contexts on performance and learning outcomes. Journal of Educational Psychology, 105(4), 932–945.
Read the paper (opens in a new tab)
In HeroMemo: lessons follow each child's age, level and answers; parents and teachers can move the level up or down; examples come from the child's favourite world.
Stories, with your child as the hero
We remember stories better than plain explanations, and we remember best what is about ourselves.
Stories were understood more easily and recalled better than explanatory texts, and this result held across many kinds of studies.
What was measured: Meta-analysis of more than 75 samples and over 33,000 participants.
Mar, Li, Nguyen & Ta (2021). Memory and comprehension of narrative versus expository texts: A meta-analysis. Psychonomic Bulletin & Review, 28(3), 732–749.
Read the paper (opens in a new tab)People who linked a list of words into a story recalled 93% of them; people who studied the usual way recalled 13%.
What was measured: Twelve word lists, college students: a list-learning task, not a full curriculum.
Bower & Clark (1969). Narrative stories as mediators for serial learning. Psychonomic Science, 14(4), 181–182.
Read the paper (opens in a new tab)Information we connect to ourselves is remembered better than the same information thought about in other ways.
What was measured: Meta-analysis of memory experiments on the self-reference effect.
Symons & Johnson (1997). The self-reference effect in memory: A meta-analysis. Psychological Bulletin, 121(3), 371–394.
Read the paper (opens in a new tab)
In HeroMemo: every lesson is a short story, and your child is the hero.
Answering questions beats reading again
Trying to remember an answer makes a memory stronger than reading the same page again, in labs and in real classrooms.
Students who tested themselves after reading remembered 61% a week later; students who re-read the text remembered 40%.
What was measured: Prose passages, college students, recall after 5 minutes and after a week.
Roediger & Karpicke (2006). Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention. Psychological Science, 17(3), 249–255.
Read the paper (opens in a new tab)Practice tests were more helpful for learning than re-studying and every other comparison method.
What was measured: Meta-analysis of experiments on practice testing, across ages, subjects and test types.
Adesope, Trevisan & Sundararajan (2017). Rethinking the Use of Tests: A Meta-Analysis of Practice Testing. Review of Educational Research, 87(3), 659–701.
Read the paper (opens in a new tab)In real schools and classrooms, from elementary school to university, retrieval practice consistently improved learning.
What was measured: Systematic review of classroom studies.
Agarwal, Nunes & Blunt (2021). Retrieval Practice Consistently Benefits Student Learning: a Systematic Review of Applied Research in Schools and Classrooms. Educational Psychology Review, 33(4), 1409–1453.
Read the paper (opens in a new tab)
In HeroMemo: the Hero challenge after each lesson, and quick memory rescues.
Reviewing just before forgetting
Much of what we learn fades within days. Reviews spread out over time make memories last far longer than cramming.
Spreading study over time beat studying all at once, across hundreds of experiments. The longer something must be remembered, the longer the best gap between reviews.
What was measured: Review and quantitative synthesis of experiments on verbal memory.
Cepeda, Pashler, Vul, Wixted & Rohrer (2006). Distributed practice in verbal recall tasks: A review and quantitative synthesis. Psychological Bulletin, 132(3), 354–380.
Read the paper (opens in a new tab)A modern repeat of Ebbinghaus's classic experiment confirmed the forgetting curve: memory drops fast in the first hours and days, then more slowly.
What was measured: One participant learning lists of syllables, retested from 20 minutes up to 31 days.
Murre & Dros (2015). Replication and Analysis of Ebbinghaus' Forgetting Curve. PLOS ONE, 10(7), e0120644.
Read the paper (opens in a new tab)Of ten common study techniques, practice testing and spaced practice were rated the most useful. Memory tricks on their own (keywords and imagery) were rated less useful, because their benefits are narrower.
What was measured: A review of research across ages, materials and settings.
Dunlosky, Rawson, Marsh, Nathan & Willingham (2013). Improving Students' Learning With Effective Learning Techniques: Promising Directions From Cognitive and Educational Psychology. Psychological Science in the Public Interest, 14(1), 4–58.
Read the paper (opens in a new tab)
In HeroMemo: each idea has its own memory curve, and a review comes back just as it starts to fade.
Games that teach
Well-designed learning games help children learn and remember more than ordinary lessons, on average.
Serious games led to more learning (0.29 standard deviations) and better retention (0.36) than conventional teaching, but they were not found to be more motivating.
What was measured: Meta-analysis of experiments comparing games with conventional instruction.
Wouters, van Nimwegen, van Oostendorp & van der Spek (2013). A meta-analysis of the cognitive and motivational effects of serious games. Journal of Educational Psychology, 105(2), 249–265.
Read the paper (opens in a new tab)Digital games improved learning compared with non-game teaching (0.33 standard deviations), and better game designs added more (0.34).
What was measured: Meta-analysis of studies with students from kindergarten to university.
Clark, Tanner-Smith & Killingsworth (2016). Digital Games, Design, and Learning: A Systematic Review and Meta-Analysis. Review of Educational Research, 86(1), 79–122.
Read the paper (opens in a new tab)
In HeroMemo: short games and quests built only from the lesson's checked facts.
Pictures together with words
Words and pictures together are remembered better than words alone, especially when a voice explains the picture.
Words and images are stored as two linked memory codes, so information learned both ways is remembered better.
What was measured: A review of decades of research on dual coding.
Paivio (1991). Dual coding theory: Retrospect and current status. Canadian Journal of Psychology, 45(3), 255–287.
Read the paper (opens in a new tab)Learning improves when a voice explains the pictures instead of on-screen text, and when distracting extras are left out.
What was measured: A summary of experiments on learning with pictures and words.
Mayer & Moreno (2003). Nine Ways to Reduce Cognitive Load in Multimedia Learning. Educational Psychologist, 38(1), 43–52.
Read the paper (opens in a new tab)
In HeroMemo: every key idea has a picture, narration reads it aloud, and labels are real text.
Memory tricks
Memory tricks such as links, images and little stories help children remember facts and new words.
Twenty years of classroom studies showed that memory strategies reliably help children remember facts.
What was measured: A review of classroom research on mnemonic strategies.
Levin (1993). Mnemonic Strategies and Classroom Learning: A Twenty-Year Report Card. The Elementary School Journal, 94(2), 235–244.
Read the paper (opens in a new tab)Mnemonics help students remember facts and vocabulary, and they work best together with understanding.
What was measured: A review of research on mnemonics in education.
Putnam (2015). Mnemonics in education: Current research and applications. Translational Issues in Psychological Science, 1(2), 130–139.
Read the paper (opens in a new tab)After six weeks of memory-palace training, ordinary people recalled 62 of 72 words, up from 26, and their brain networks became more like those of memory champions.
What was measured: Adults: word-list recall before and after training, with brain scans.
Dresler, Shirer, Konrad et al. (2017). Mnemonic Training Reshapes Brain Networks to Support Superior Memory. Neuron, 93(5), 1227–1235.
Read the paper (opens in a new tab)
In HeroMemo: a memorable hook for each key idea, always paired with questions and spaced review.
Curiosity and questions
When we are curious, we learn more, and we remember it better.
When people were curious about a question, they remembered its answer better a day later, and even remembered unrelated things they saw while curious.
What was measured: Adults answering trivia questions, with brain scans.
Gruber, Gelman & Ranganath (2014). States of Curiosity Modulate Hippocampus-Dependent Learning via the Dopaminergic Circuit. Neuron, 84(2), 486–496.
Read the paper (opens in a new tab)
In HeroMemo: in “Ask Momo”, your child taps a question about the slide and hears the answer, the slide's checked fact.
What the research does not say
- These are average results from many studies. Every child is different.
- These studies were not about HeroMemo. Each one supports one part of how HeroMemo works.
- The famous 'two sigma' result came from small studies; larger reviews found smaller, but still large, gains.
- Games help learning on average, but they are not automatically more motivating.
- Memory tricks work best for facts and new words. That is why HeroMemo always pairs them with questions and spaced review.
We are preparing a pilot study with schools to measure HeroMemo itself: a test before and after, then again one week and one month later.
