Dale's Pyramid: where the percentages in the learning pyramid come from

Dale's Pyramid: looking for the original study behind the percentages

"Dale's pyramid" claims to put numbers on what we retain depending on the teaching method, from 5% for a lecture to 90% for teaching others. Edgar Dale never published these figures: his cone of experience (1946) ranked experiences from concrete to abstract, with no percentages. No original study has been found.

Hello to you, the movement professional,

You have certainly seen this slide go by in a training course: a colorful pyramid, figures stepping up. We supposedly retain 5% of a lecture, 10% of what we read, 20% of a video and 30% of a demonstration. We would climb to 50% by discussing in a group, 75% by practicing and 90% by teaching others. At the bottom, in small print: Edgar Dale.

Before reading on, do the exercise: look for the original study, the one that would have measured these seven figures, with its participants and the delay between the lesson and the test. You will find hundreds of reuses, and not a single article that describes the experiment.

If you are looking for the pyramid of the vestibular system and the body schema, that is the pyramid of learning by Taylor and Trott, taken up by Williams and Shellenberger. Here, we follow the trail of the percentages, then keep what is useful in session.

Look at the figures with a coach's eye. They land on round values and move neither with the audience nor with the delay before the test. You would reject a 1RM announced without the load or the athlete. And the man whose name appears at the bottom of the slide wrote none of these figures.

Edgar Dale's cone of experience, from concrete to abstract

Edgar Dale, a professor at Ohio State University, published Audio-Visual Methods in Teaching in 1946. In it he proposes the cone of experience, revised in 1954 and then in 1969, which ranks learning experiences by their degree of abstraction.

At the base, direct experience, the kind you see, handle and feel. Moving up come contrived experience, dramatized experience, demonstration, field trip, exhibit, motion picture, radio and still picture. At the top, visual symbols, then words. Dale illustrated it this way: the word "horse" written on a page has neither the gait nor the sound of a horse. The higher you go, the more the learner watches instead of taking part.

In 1969, Dale himself described his cone as a "visual analogy". He even warns against too much of the concrete, which can get in the way of generalization, the ability to draw a rule from several cases. For him, good teaching combines the levels.

Where the percentages of the learning pyramid come from

Two investigations sorted it out. In 2014, Deepak Subramony, Michael Molenda, Anthony Betrus and Will Thalheimer devoted a special issue of Educational Technology to the "corrupted cone". In 2018, Kåre Letrud and Sigbjørn Hernes combed through digital libraries. Their results untangle three stories glued to one another.

The oldest is a maxim, of which Letrud and Hernes find early versions as far back as the 1850s, before the psychology of learning even existed. It therefore cannot come out of a laboratory.

The second goes through corporate training. In 1967, D. G. Treichler, an employee of the oil company Mobil, signed in an audiovisual journal the most widely reused numerical version. In it we supposedly retain 10% of what we read and up to 90% of what we say while doing, on the strength of studies never referenced. For Michael Molenda, this alleged data looks fabricated by oil-company trainers in the 1960s.

The third name often cited on the slide: the National Training Laboratories, in Bethel, Maine. The institute says it developed its pyramid in the early 1960s, but has never been able to supply the research supposed to underpin it, which is reportedly lost.

Nobody knows who put the figures onto the shape of the cone. The oldest trace found by Subramony and his colleagues dates from the late 1970s, in a train-the-trainer document. Since then, the montage has traveled from slide to slide, with an academic's name as a guarantee.

The pyramid's percentages: what those who checked found

In 2007, in the journal Education, J. P. Lalley and R. H. Miller put the pyramid through the sieve and found no credible research behind the percentages. Method by method, the literature shows them that each helps retention in certain contexts, without any of them winning every time. Reading and direct instruction, relegated to the top of the pyramid, even seem to them undervalued. Five years later, Kåre Letrud hammers the point home in the same journal and asks the NTL to withdraw its diagram.

In medical education, a discipline nonetheless steeped in critical reading, Ken Masters counted in 2013 no fewer than 43 peer-reviewed publications citing the pyramid. The sources are misread or circular, and the figures change from one article to the next. He recommends no longer referring to it.

When he repeats the search for 2012 to 2018, he finds 41 documents, rising citations, and authors who cite the rebuttals themselves in support of the pyramid. His 2020 conclusion, translated without softening it: the pyramid is worthless, its statistics too, and they do not come from Edgar Dale. If people trained to read studies fall for it, a well-made slide deserves our wariness too.

The level of evidence, without rounding in either direction

Once the figures have fallen, we would be tempted to throw out the intuition that made them plausible as well, when part of it holds up very well against the studies.

No universal percentage in the right-hand column. A round figure is remembered better than a nuance, hence these stubborn beliefs about learning that resist so well.

What research supports for learning: active learning and the testing effect

Scott Freeman and his colleagues (2014) compiled 225 studies comparing lecturing with active learning in science, at university. With active learning, exam scores gain about 6% and the failure rate drops from 33.8% to 21.8%. The gap matters, but it is a long way from the slide's 5% against 90%.

Henry Roediger and Jeffrey Karpicke (2006) had students read texts, then either reread them or test themselves in writing, with no feedback at all. At five minutes, rereading wins. At two days and at one week, the ranking reverses. Those who reread nonetheless felt more confident. With a winner that changes depending on the delay, a fixed percentage per method no longer makes any sense.

That leaves the top of the slide. Aloysius Koh and his colleagues (2018) separated two things that are confused, teaching and recalling. One week later, those who taught without their notes retain better than those who had them in front of their eyes, and simple recall, without teaching anything, also gives this advantage.

Teaching a movement: what motor learning adds

You teach movements, and motor learning has its own cousin of the testing effect.

John Shea and Robyn Morgan (1979) had people learn three fast sequences of arm movements, in blocks or in a random order. During practice, blocks do better. At ten minutes and then at ten days, random order retains better and transfers better to a new task. Timothy Lee and Richard Magill (1983) explained it this way: switching tasks on every trial makes you forget part of the action plan, which has to be rebuilt, and therefore retrieved from memory.

The session that looks cleanest therefore sometimes leaves the fewest traces. That is the whole gap between immediate performance and learning, and the reason a movement is judged on retention and transfer rather than on the last trial.

The effect has its limits. The meta-analysis by Stanisław Czyż and colleagues (2024), covering 34 studies, finds a benefit of random order on transfer. Clear in the laboratory, it becomes weak and non-significant in the field, almost nil in young people, and is confirmed in adults. The order of variants is therefore dosed according to the audience, like all practice variability.

The pyramid of learning or Dale's pyramid: if someone talks to you about the vestibular system, it's the other one

When someone talks to you about the vestibular system, body schema or motor planning, they mean the pyramid of learning by Kathleen Taylor and Maryann Trott (1991). It is built on Ayres' sensory integration and was popularized in 1996 by Mary Sue Williams and Sherry Shellenberger.

Its five levels run from the central nervous system to cognition, with the vestibular system among the senses of the second. No retention rate appears in it, and its strict hierarchy has not been demonstrated. We detail it in our article on Williams and Shellenberger.

In training or in session: what to say instead of the percentages

When the slide shows up in a training course, there is no need to put it on trial. One sentence is enough: "These figures never had a study behind them, and Edgar Dale did not write them. The testing effect, on the other hand, rests on dozens of experiments." Then you bring it to life instead of projecting it.

With an athlete, once the movement is roughed out, mix the order of the variants instead of chaining ten identical repetitions, even if it means living through a slightly less clean session. Learning is judged a few days later, with no cue just before, following the logic of test, stimulation, retest that we also apply in Neuro-Postural Reprogramming.

With a child or a complete beginner, still at the first of the Fitts and Posner stages, our practitioner's hypothesis is to keep more blocks at the start. Czyż's meta-analysis finds an almost nil effect in young people, and nothing yet shows that it is strong in beginners.

The percentages were hiding a question: with equal method, what makes learning more costly for a given athlete or child? A good part of the answer is played out lower down, in the sensory and motor foundations described by the Taylor and Trott pyramid.

What remains of a session is what the athlete had to go and retrieve on their own.

👉 I want to intervene in the right place.

Frequently asked questions about Dale's pyramid

What is Dale's pyramid?

It is the common name for a diagram that assigns a retention rate to each teaching method, from 5% for a lecture to 90% for teaching others. Falsely attributed to Edgar Dale and spread by the National Training Laboratories, it rests on no study that has been found.

Did Edgar Dale invent the percentages of the learning pyramid?

No. His cone of experience (1946, revised in 1954 and 1969) ranks experiences from concrete to abstract without a single figure. The most widely reused numerical version appeared in 1967 in an audiovisual journal, from the pen of a Mobil employee, and the oldest known trace of the figures placed on the cone dates from the late 1970s.

Is it true that we retain 10% of what we read?

No study shows it. What you retain from a reading depends on the content, the reader, the delay and what you do afterward. Testing yourself on a text leaves, a few days later, clearly more traces than rereading it, as Roediger and Karpicke showed in 2006.

What is Dale's learning cone?

Its exact name is the cone of experience. In it, Edgar Dale ranks learning experiences, from direct experience at the base up to words at the top. He presented it as a visual analogy and advised combining concrete and abstract experiences.

What is the difference between Dale's pyramid and the pyramid of learning?

Dale's pyramid lines up percentages with no source. The pyramid of learning by Taylor and Trott (1991), spread by Williams and Shellenberger (1996), describes in five levels the sensory and motor foundations of learning. It is presented in the article on the pyramid of learning.

My child retains little of what they are taught despite good methods: who should I turn to?

When the difficulty lasts and weighs on schooling or daily life, the assessment belongs to the doctor who follows the child, the occupational therapist or the psychomotor therapist. The movement professional observes and refers, then works on movement. LabO RNP is neither medical nor paramedical, and comes in addition to a course of care, never in its place.

Sources

Dale E. (1946). Audio-Visual Methods in Teaching. New York: Dryden Press. Revised editions in 1954 and 1969.

Subramony D.P., Molenda M., Betrus A.K., Thalheimer W. (2014). The mythical retention chart and the corruption of Dale's Cone of Experience. Educational Technology, 54(6), 6-16.

Letrud K., Hernes S. (2018). Excavating the origins of the learning pyramid myths. Cogent Education, 5(1), 1518638.

Molenda M. (2003). Cone of Experience. In A. Kovalchick, K. Dawson (eds.), Education and Technology: An Encyclopedia. Santa Barbara: ABC-CLIO.

Lalley J.P., Miller R.H. (2007). The learning pyramid: Does it point teachers in the right direction? Education, 128(1), 64-79.

Letrud K. (2012). A rebuttal of NTL Institute's learning pyramid. Education, 133(1), 117-124.

Masters K. (2013). Edgar Dale's Pyramid of Learning in medical education: a literature review. Medical Teacher, 35(11), e1584-e1593.

Masters K. (2020). Edgar Dale's Pyramid of Learning in medical education: Further expansion of the myth. Medical Education, 54(1), 22-32.

Freeman S. et al. (2014). Active learning increases student performance in science, engineering, and mathematics. PNAS, 111(23), 8410-8415.

Roediger H.L., Karpicke J.D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249-255.

Rowland C.A. (2014). The effect of testing versus restudy on retention: A meta-analytic review of the testing effect. Psychological Bulletin, 140(6), 1432-1463.

Koh A.W.L., Lee S.C., Lim S.W.H. (2018). The learning benefits of teaching: A retrieval practice hypothesis. Applied Cognitive Psychology, 32(3), 401-410.

Shea J.B., Morgan R.L. (1979). Contextual interference effects on the acquisition, retention, and transfer of a motor skill. Journal of Experimental Psychology: Human Learning and Memory, 5(2), 179-187.

Lee T.D., Magill R.A. (1983). The locus of contextual interference in motor-skill acquisition. Journal of Experimental Psychology: Learning, Memory, and Cognition, 9(4), 730-746.

Czyż S.H., Wójcik A.M., Solarská P. (2024). The effect of contextual interference on transfer in motor learning: a systematic review and meta-analysis. Frontiers in Psychology, 15, 1377122.

By the LabO RNP team