Tracing the "Learning Pyramid" and "Cone of Learning" Percentages
In corporate training, professional development, and educational presentations, presenters frequently display a diagram called the "Learning Pyramid" or "Cone of Learning." The slide claims that people remember:
- 10% of what they read
- 20% of what they hear
- 30% of what they see
- 50% of what they see and hear
- 70% of what they say
- 90% of what they say and do
These incredibly neat, tidy multiples of ten are widely taught as established cognitive science. However, educational researchers have thoroughly debunked these percentages as complete fabrications. The true history of this graphic is a cautionary tale of how a theoretical model of abstraction was corrupted and merged with made-up numbers.
The True Origin: Edgar Dale's Cone of Experience
In his 1946 book Audio-Visual Methods in Teaching, Edgar Dale (an expert in audiovisual education) introduced the "Cone of Experience."
- No Percentages: Dale's original diagram contained absolutely no percentages or numbers.
- A Model of Abstraction, Not Retention: The Cone was not a model of memory retention or learning effectiveness. Instead, it was a visual metaphor for levels of mental abstraction. "Words" (verbal symbols) were at the top of the cone because they are the most abstract way to represent reality. "Direct, purposeful experiences" (real-life action) were at the base because they are the most concrete.
- No Hierarchy of Value: Dale explicitly argued that no level of the cone was inherently superior to another. A teacher should use different modalities depending on the context, and abstract verbal learning is highly efficient for learners who already have concrete background knowledge. In fact, by the book's third edition in 1969, Dale added a six-page chapter titled "Some Possible Misconceptions" to warn readers against interpreting the Cone as a hierarchy of learning effectiveness.
How the Percentages Were Grafted On
If Dale's original Cone had no numbers, where did the 10/20/30/50/70/90% figures come from?
-
The 1967 Mobil Oil Article: In 1967, an employee of Mobil Oil Company (then Socony Mobil Oil) named D. G. Treichler published an article in a magazine called Film and Audio-Visual Communications titled "Are You Missing the Boat in Training Aids?" Treichler wrote:
"People generally remember: 10% of what they read, 20% of what they hear, 30% of what they see, 50% of what they hear and see..."
Treichler provided no citations, data, or empirical evidence for these numbers. Researchers believe he may have adopted them from casual, unscientific training materials circulated during World War II at the military's Aberdeen Proving Ground, but no research has ever been found to support them.
-
The Amalgamation: Sometime in the late 1960s or 1970s, an unknown designer took Treichler's fabricated percentages and literally pasted them onto Edgar Dale's Cone of Experience. Because Dale's name was highly respected, the newly minted "Cone of Learning" or "Learning Pyramid" suddenly possessed immense academic authority.
-
The "Lost Data" of the NTL Institute: The National Training Laboratories (NTL) Institute for Applied Behavioral Science became a major propagator of the Learning Pyramid. When researchers like Dr. Will Thalheimer and Deepak Prem Subramony repeatedly pressured the NTL for the original empirical studies backing up the pyramid, the NTL responded that they knew the research had been conducted in the 1960s but they had "lost the original data."
Why the Learning Pyramid is a Myth
Cognitive scientists and educational psychologists (such as Daniel Willingham) have pointed out several fundamental flaws in the Learning Pyramid:
- Tidy Multiples of 10: Real scientific data is messy. It never results in perfectly neat increments of 10% (10, 20, 30, 50, 70, 90).
- Quality Over Modality: The effectiveness of learning does not depend on the sensory channel (reading vs. seeing), but on the cognitive effort and quality of processing. An active peer-coaching exercise can be highly confusing if the students don't understand the material, while reading a well-written book can lead to profound, long-term comprehension.
- Context is Ignored: Retention rates vary wildly depending on the difficulty of the material, the learner's prior knowledge, their motivation, and how memory is tested (e.g., immediate recall vs. a delayed test).