5 surprising insights from the science of successful learning
The work of Reda Sadki offers a provocative, often counter-intuitive critique of how we learn, lead, and solve complex problems.
Here are five surprising insights from his body of work.
1. Text is superior to video for learning
In an era where educational technology is obsessed with video content, immersive simulations, and flashy multimedia, Sadki argues for the humble written word.
He asserts that the push for multimedia is often a “deception” that confuses engagement with entertainment.
In Richard Mayer’s research on multimedia for learning actually proves text works better, Sadki re-examines the foundational science of instructional design.
He points out that multimedia often creates “cognitive waste” by forcing the brain to split attention between visual and auditory streams.
He argues that well-structured text is “cognitively quiet” and far better suited for the high-level critical thinking required in complex fields.
He doubles down on this in Against chocolate-covered broccoli: text-based alternatives to expensive multimedia content.
Here, he describes multimedia as an economic dead end.
He argues that text is not only cheaper and easier to update but also creates a more equitable learning environment for professionals in low-bandwidth settings.
2. Gamification is a “disaster” for humanitarian learning
While many organizations rush to “gamify” learning with badges, points, and leaderboards, Sadki calls this trend a “dead end.”
He argues that gamification is simply “lipstick on the pig of behaviorism,” a discredited theory that treats learners like rats in a maze responding to stimuli.
In Why gamification is a disaster for humanitarian learning, he makes a blistering case that games fail to model the complexity of the real world.
He points out that the dominant culture of video games often relies on violence and competition, which are antithetical to humanitarian values.
He argues that professionals facing life-and-death decisions need critical reasoning skills, not the artificial dopamine hits of a game.
3. Low completion rates can be a sign of success, not failure
In the world of online courses, a low completion rate is usually seen as a failure of design.
Sadki flips this metric on its head.
He suggests that in professional settings, “completion” is a vanity metric, part of the legacy of education systems that kept learners in closed environments.
In Online learning completion rates in context: Rethinking success in digital learning networks, he argues that busy professionals often engage with learning to solve a specific problem.
Once they find the solution, they leave.
This “drop-off” is actually efficient learning in action.
He warns that optimizing for completion often leads to dumbing down content rather than increasing its impact.
4. The “transparency paradox”: health workers are using AI in secret
One of Sadki’s most startling recent observations comes from his work with frontline health workers.
He reveals that professionals in the Global South are already using advanced Artificial Intelligence (AI) tools, but they are forced to hide this fact.
In Artificial intelligence, accountability, and authenticity: knowledge production and power in global health crisis, he describes a “transparency paradox.”
Global health systems are often punitive.
If a health worker admits to using AI to help draft a report or analyze data, their work is devalued as “inauthentic,” even if the quality is higher.
This forces innovation underground and prevents organizations from learning how to effectively partner with AI.
He expands on the solution in A global health framework for Artificial Intelligence as co-worker to support networked learning and local action, arguing that we must legitimize AI as a “co-worker” rather than a cheat.
5. Cascade training is mathematically doomed to fail
Finally, Sadki uses simple mathematics to dismantle one of the most common methods of training in the world: the “cascade” model, where experts train trainers, who train others.
In Why does cascade training fail?, he demonstrates that information loss at every level of the cascade is inevitable.
He argues that this model persists not because it works, but because it is convenient for hierarchical organizations.
He offers a stark alternative in Calculating the relative effectiveness of expert coaching, peer learning, and cascade training, where he proves that peer learning networks are the only model capable of scaling without losing quality.
#ArtificialIntelligence #completionRates #gamification #globalHealth #learningStrategy #multimediaLearning #RichardMayer #TheGenevaLearningFoundation
