Why does Kubrio teach through projects instead of lessons?
Kubrio is built around projects because that is the direction the evidence points. The largest analysis of learning research ever done — 225 studies — found students in listen-only classrooms were about one and a half times more likely to fail than students who learned actively. Kids learn what they practice, so Kubrio makes the practice the unit.
Kubrio hands your kid a project instead of a lesson. Parents reasonably want to know whether that is a real bet on the evidence or just a nicer-sounding way to spend an afternoon. Here is the research, including its limits.
The biggest study in the field. In 2014 a team pooled 225 studies comparing traditional lecturing with active learning in undergraduate science, engineering, and mathematics courses. Two numbers came out of it. Average exam scores rose by roughly half a letter grade when students learned actively. And failure rates diverged sharply: students in traditional lecture courses were about 1.5 times more likely to fail than students in the active-learning versions of the same courses (Freeman et al., 2014). It remains the largest analysis of its kind, and it is the single result we lean on most.
Not all "doing" is equal. The distinction that matters is not activity versus stillness — it is how much the kid has to generate. The ICAP framework sorts learning behaviors into four modes and ranks them: Interactive (building something in dialogue with someone else) beats Constructive (producing something that goes beyond what was given — an explanation, a design, a drawing) beats Active (manipulating the material, like highlighting or repeating) beats Passive (watching or listening) (Chi & Wylie, 2014). That ordering is the reason "make a manga about it" is a categorically different activity from "watch a video about it." Making forces the kid to produce something nobody handed them.
Projects, specifically. Active learning is a wide family of methods, so the narrower question is whether carrying a whole project to completion holds up. A meta-analysis comparing project-based learning with traditional instruction found a positive effect on students' academic achievement, in the medium-to-large range (Chen & Yang, 2019). Stated plainly: kids who learn by finishing a project generally do better on the same measures than kids taught the same material directly.
The honest fine print. Most of this evidence comes from classrooms, and much of it from older students — Freeman's 225 studies are undergraduate courses, not nine-year-olds making a film at the kitchen table. Meta-analyses also pool very different interventions under one label, so "active learning" and "project-based learning" name families of methods, not a single recipe anyone can hand you. And every result here is an average across thousands of students; an average is not a promise about your kid. What the literature supports is a direction, not a guarantee: learning by doing is the best-evidenced bet in the field, and we would rather say that than claim more.
What Kubrio does with it. We make the project the unit of learning instead of the reward at the end of one: the shelf is always open, and your kid picks what pulls them rather than working through an assigned order. The AI crew sits in the Interactive corner of that ranking on purpose — it asks, reacts, and pushes for a finished result, and it never makes the thing for them. And every project ends published rather than saved, because a real finish line is what makes the making worth doing twice.
Sources.
- Freeman, S., et al. (2014). Active Learning Increases Student Performance in Science, Engineering, and Mathematics. Proceedings of the National Academy of Sciences, 111(23), 8410–8415.
- Chi, M. T. H., & Wylie, R. (2014). The ICAP Framework: Linking Cognitive Engagement to Active Learning Outcomes. Educational Psychologist, 49(4), 219–243.
- Chen, C.-H., & Yang, Y.-C. (2019). Revisiting the Effects of Project-Based Learning on Students' Academic Achievement: A Meta-Analysis Investigating Moderators. Educational Research Review, 26, 71–81.
Related reading: the science behind Claire's talking prompts — what the research says about how to respond once your kid has actually made something.
Frequently asked questions
Is "learning by doing" actually proven?
It is the best-supported direction in learning research, which is not the same as proven for any one kid. The strongest single piece of evidence is a meta-analysis of 225 studies finding higher exam scores and substantially lower failure rates under active learning than under traditional lecturing (Freeman et al., 2014), and a separate meta-analysis found project-based learning outperformed traditional instruction on academic achievement (Chen & Yang, 2019). Those are averages across large numbers of students, not a guarantee about yours.
Do kids miss the fundamentals if they only make things?
That is the fair worry, and it is the thing these studies actually measured. Freeman's comparison was on exam performance in the same courses, and Chen and Yang's was on academic achievement — in both cases the active and project-based groups came out ahead, not behind. The evidence does not suggest making things comes at the cost of the content; it suggests making things is how the content sticks.
Does this research cover kids my kid's age?
Only partly, and we would rather say so. The 225-study meta-analysis is undergraduate coursework, so it tells us about the mechanism, not about eight-year-olds. Project-based learning has been studied across a wider span of school ages, but none of it was run on your kid in your house. We treat all of it as the best available direction and then watch what your kid actually finishes.