Duolingo uses a data-driven approach to language education, and relies on user metrics to customize the learning experience for each user.
Traditional learning vs Duolingo[edit | edit source]
Contrary to traditional learning, Duolingo uses its data to improve the metrics and restructure the learning experience for all learners. Due to the massive number of learners in Duolingo, this makes it uniquely positioned to identify learning difficulties and create interventions to help users overcome them. It also uses A/B testing to evaluate possible interventions that can benefit students.
Translation and Duolingo[edit | edit source]
Most Duolingo exercises rely on translation from and into the target language, though some of its exercises are based on speaking or listening to sentences in the target language. These latter features are not available for community-created languages created in the Incubator.
Language acquisition theory[edit | edit source]
Duolingo is based on cognitive science and educational psychology research. In particular it uses active recall (the "testing effect"), which relies on complete sentences that have to be generated from memory, as well as Spaced repetition (the "forgetting curve"), adaptive and personalized to users and based on the mathematical models.
Courses can be based on core vocabulary (word frequency), which consists of teaching frequent words first while leaving less common words for later, and makes use of multi-modal learning, a combination of senses in every lesson (e.g., listening, reading) to help learn the ideas consciously and subconsciously.
Research[edit | edit source]
- Existing research indicates that Duolingo is more effective than Rosetta stone.
Criticism[edit | edit source]
- Some claim that when using Duolingo "you are the product".