The Science Behind
MathKix

Hermann Ebbinghaus discovered in 1885 that memories decay exponentially - unless they are reviewed at strategically increasing intervals. Paul Pimsleur later formalized graduated-interval recall, showing that each successful review pushes the next optimal review further into the future.

• Implements the SM-2 algorithm - the same system used by millions of learners worldwide
• Child-friendly modification: the ease factor never decreases on failure, preventing compounding difficulty for struggling learners
• Review intervals start at 1-3 days, then 6 days, then multiply by each standard's personal ease factor
• Every Common Core standard is tracked independently with full score history

Lev Vygotsky proposed in 1978 that learning happens most effectively in the zone of proximal development - the space between what a child can do independently and what they can achieve with guidance. Material that is too easy leads to boredom; material that is too hard leads to frustration.

• Scores every individual question from the entire grade-level pool across 6 factors: mastery need, spaced repetition urgency, difficulty fit, domain weight, topic affinity, and variety
• Each session mixes domains at different difficulties: easy questions for struggling standards, hard questions for strong ones - all in the same sitting
• Difficulty targeting is per-standard: a child who struggles with geometry gets difficulty-1 questions while simultaneously receiving difficulty-3 arithmetic to stay sharp
• Mastery level 3 (full mastery) is only reachable by consistently solving hard questions - easier questions build toward it but cannot skip the final step

Carol Dweck's research demonstrated that praising effort and strategy - rather than innate ability - leads to greater persistence, resilience, and achievement in mathematics. Children with a growth mindset treat mistakes as learning opportunities rather than evidence of failure.

• Language adapts to grade level: Grade 1 gets 1-2 sentence, 50-word responses with concrete analogies; Grade 5 gets full mathematical vocabulary
• Four-tier scaffolding: Explore (ask a question) - Nudge (small hint) - Scaffold (reveal one fact) - Direct (walk through a step)
• Emotion-specific responses: detects frustration, confusion, excitement, and disengagement - then adjusts tone and approach
• Never says "wrong" - instead names the specific strategy used, reframes mistakes as progress, and asks guiding questions

Mihaly Csikszentmihalyi's flow research showed that optimal engagement occurs when challenge closely matches ability. When the balance tips toward frustration, learners disengage. The key is to detect the shift and adapt in real time - before the child gives up.

• Monitors response time, error streaks, and session duration in a sliding 5-question window
• Detects slowing (response > 2x average), error streaks (3+ wrong), fatigue (20+ minutes), and disengagement
• Difficulty and topic adapt before frustration - not after the child has already checked out
• Topic affinity system with 14-day half-life tracks which domains naturally engage each child, weighting future lessons toward intrinsic motivation