These six principles from learning science guide how we teach.
Click on each point on the hexagon to discover the research-backed learning strategies that power Grassroot's AI tutoring system.
Working memory holds only a few items at once. When lessons are cluttered, mental effort gets wasted on confusion instead of learning.
We do: One idea at a time. No clutter.
Sweller, J. (2023). The Development of Cognitive Load Theory. Educational Psychology Review, 35(4)
Testing isn't just assessment—it's learning. Students who tested themselves retained twice as much as those who restudied.
We do: Regular recall prompts, not re-explanations.
Roediger, H.L. III, & Butler, A.C. (2011). The critical role of retrieval practice in long-term retention. Trends in Cognitive Sciences, 15(1), 20-27
Ericsson found elite violinists accumulated 10,000+ hours by age 20—5,000 more than average performers. The difference wasn't talent; it was structured practice targeting weaknesses with immediate feedback.
We do: Find your weak spots. Focus there.
Ericsson, K.A., Krampe, R.T., & Tesch-Römer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review, 100(3), 363-406
Instruction that helps beginners can actually harm experts. Novices need step-by-step guidance; experts find the same guidance distracting.
We do: Track your level. Adjust accordingly.
Kalyuga, S., et al. (2003). The Expertise Reversal Effect. Educational Psychologist, 38(1), 23-31
Students who generate their own explanations understand more than those who passively receive them. Best sequence: attempt your own representation first, then view the expert version as feedback.
We do: You explain first. Then we show ours.
Fiorella, L. (2023). Making Sense of Generative Learning. Educational Psychology Review
Metacognition is knowing when you understand and when you're confused. Students with this awareness self-correct; those without stay stuck.
We do: Prompt you to rate confidence and flag confusion.
Fiorella, L., & Mayer, R.E. (2021). Fostering Metacognition to Support Student Learning and Performance. CBE—Life Sciences Education
students fail
Traditional lecture-based learning has a 33% failure rate in STEM courses
Freeman et al. (2014), PNAS
better outcomes
Students move from 50th to 85th percentile with personalized AI tutoring
Wang et al. (2024), Educational Computing Research