The research behind every mission
Cadenza brings together three lines of language research. Here is what each one says, how strong the evidence is, and where questions are still open.
Prosodic shadowing: the music first
Speaking along with a native voice, a quarter of a second behind, while paying attention to melody rather than meaning.
In a traditional listen-and-repeat drill, the learner hears a sentence, stops, remembers it, often translates it, and then says it. Shadowing is different: listening and speaking happen at the same time.
Because the student trails the native voice by only 250–500 milliseconds, there's no time to translate. Sound goes straight from hearing to speaking, and the student's timing locks onto the rhythm of the sentence.
Prosodic shadowing narrows the focus further, to pitch, stress and rhythm. That makes it achievable for learners who can't yet copy every word.
What happens in the brain
Melody before words
Researchers teach shadowing in three steps, from easiest to hardest. Cadenza's Echo step sits in the middle. The Watch step before it makes sure students already understand most of what they hear, because shadowing works best on material learners mostly understand.
- 1
Mumble
Copy the sound and rhythm only.
- The Echo step2
Melody
Copy pitch, stress and rhythm, without worrying about every word.
- 3
Words and meaning
Copy every word while following what it means.
The 250–500 ms window
Shadowing studies going back to the 1950s put the limit of speech reaction near a quarter of a second. Past half a second, learners fall back on memory.
Source: Speech shadowingA direct line from ear to mouth
Speaking while listening uses the brain's dorsal stream, which maps sounds onto mouth movements. Brain scans of shadowing show hearing and speech areas working together.
Hickok & Poeppel 2007; NeuroImage, 2009Measured listening gains
In a study of 43 university English learners, nine shadowing lessons improved how well they heard individual sounds, at both levels tested.
Hamada, Language Teaching Research, 2016Sixty seconds, then stop
Shadowing is intense, and learners report tiring in long sessions. Each Echo loops a short clip for one minute rather than pushing through a long one.
Asian-Pacific Journal of Second and Foreign Language Education, 2018Guiding by example
How bilingual families teach, and what classroom research says about feedback during conversation.
One person, one language is a strategy bilingual families have used for over a century: each parent sticks to one language with the child. The linguist Maurice Grammont described it in 1902. Studies of these families point to consistency as what matters most.
When the child makes a mistake, a parent typically doesn't stop to explain a rule. They recast: say the sentence back correctly and carry on. Cadenza's partner works the same way, in the language the student is learning.
A review of 15 classroom studies with 827 learners found that spoken feedback helped students improve, the gains lasted, and younger learners benefited more than older ones.
A recast
The partner says it back the right way and keeps the conversation going.
Strength: the conversation never stops, so students keep speaking.
A prompt
The partner asks the student to have another go.
Why we leave it out: prompts make students notice the gap, and in the same review they had larger effects. But each one stops the conversation, so Cadenza's partner always recasts. Teachers can still use prompts in class.
Motivation and calm
Why children who choose their work, and don't fear mistakes, keep talking.
Self-determination theory, developed by Edward Deci and Richard Ryan, holds that motivation from inside lasts when three needs are met: autonomy, competence and relatedness.
Anxiety pulls the other way. Stephen Krashen called it the "affective filter": an anxious learner takes in less. The idea is influential and still debated, but the link between anxiety and results is well measured.
Autonomy
Feeling that the choice is yours
Students pick their own projects, so every mission serves something they decided to build.
Competence
Seeing yourself get better
The ability map shows what a student can now do, mission by mission, instead of points or grades.
Relatedness
Feeling connected to others
A patient partner who listens, and classmates working on projects together in the studio.
The average link between language anxiety and achievement across 97 studies, 19,933 learners and 23 countries. The more anxious students feel, the lower their results tend to be.
Teimouri, Goetze & Plonsky, Studies in Second Language Acquisition, 2019Project-based learning
Learning a language by using it to make something.
In project-based learning, students work on an extended, real-world task and pick up skills along the way. Many progressive schools already use it for science, design and humanities.
A 2023 meta-analysis of 17 effect sizes found a moderate positive effect of project-based learning on English proficiency. The effect is real but not dramatic, which is why Cadenza combines projects with daily speaking practice rather than relying on projects alone.
Evidence at a glance
The strongest findings behind each part of Cadenza, and what kind of study they come from.
Anxiety lowers results
Meta-analysis: 97 studies, 19,933 learners. Average correlation −0.36.
Feedback in conversation works
Meta-analysis: 15 classroom studies, 827 learners. Durable effects, larger for younger learners.
Projects help moderately
Meta-analysis: 17 effect sizes for English proficiency. Moderate effect.
Shadowing sharpens listening
Controlled study: 43 university learners, nine lessons. Gains in hearing individual sounds.
Shadowing links ear and mouth
Brain imaging: hearing and speech areas work together during fast repetition.
Motivation from inside lasts
Theory with decades of research behind it: autonomy, competence and relatedness.
What we don't know yet
We'd rather you hear the limits from us.
Cadenza as a whole
Cadenza is new and hasn't been tested in an independent study yet. Each pilot measures what students can say and do, and we share the results with the school.
Age of learners
Much of the shadowing research was done with teenagers and adults learning English. Studies with younger children are still few.
Recasts versus prompts
In classroom research, prompts had larger effects than recasts. Cadenza's partner only recasts, so speaking time never stops for a correction. How much teachers add by prompting in class is a question a pilot can help answer.
References
- Hamada, Y. (2016). Shadowing: Who benefits and how? Language Teaching Research, 20(1). Link
- Hamada, Y. (2019). Shadowing: What is it? How to use it. Where will it go? RELC Journal. Link
- Hickok, G. & Poeppel, D. (2007). The cortical organization of speech processing. Nature Reviews Neuroscience, 8, 393–402.
- Auditory–motor integration during fast repetition: the neuronal correlates of shadowing. NeuroImage (2009). Link
- Motivation and attitude towards shadowing: learners' perspectives in Japanese as a foreign language. Asian-Pacific Journal of Second and Foreign Language Education (2018). Link
- Speech shadowing. Wikipedia overview
- Lyster, R. & Saito, K. (2010). Oral feedback in classroom SLA: A meta-analysis. Studies in Second Language Acquisition, 32(2), 265–302. Link
- One person, one language. Wikipedia overview
- Teimouri, Y., Goetze, J. & Plonsky, L. (2019). Second language anxiety and achievement: A meta-analysis. Studies in Second Language Acquisition, 41(2), 363–387. Link
- Ryan, R. M. & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist. Background
- Krashen, S. (1982). Principles and Practice in Second Language Acquisition. Pergamon.
- Mujiono, Weganofa, R. & Herawati, S. (2023). Project-based learning intervention and its effect on promoting English language proficiency: Meta-analysis. International Journal of Instruction, 16(4). Link