Derivita works.
The proof is in the data.
Published March 6, 2026 · The EdTech Collective & International Centre for EdTech Impact
Derivita contracted with Instructure to examine the relationship between student use of Derivita and math learning outcomes. Guided by Every Student Succeeds Act (ESSA) standards, researchers designed the study to satisfy Level III (Promising Evidence) requirements — the standard districts rely on when evaluating evidence-based edtech investments. View the complete study.
Every additional Derivita question an 8th grader answers per day is associated with a 13 percentile point improvement in their math score.
What that looks like in practice: A student scoring at the 50th percentile who answers just one more Derivita question per day would be expected to score at the 63rd percentile — moving from average to well above average with a single extra question.
in the study
answered per day
effect size (p < .001)
per question/day
A student performing at the 50th percentile would be expected to move to the 63rd percentile with just one additional Derivita question answered per day — a meaningful, measurable lift.
Researchers used a correlational design aligned with ESSA Level III evidence standards. The study was conducted during the 2024–25 school year with 1,647 eighth-grade students from a large suburban district of over 30,000 students in the Western United States.
| Measure | Description |
|---|---|
| Usage Metric | Number of Derivita questions answered per day (student level) |
| Outcome | Derivita math assessment standardized scores |
| Control Variable | Prior Derivita math assessment scores (controls for selection bias) |
| Analysis Method | Multilevel Linear Model (MLM), students nested within teachers |
This study satisfies all minimum criteria for ESSA Level III (Promising Evidence):
- ✓Correlational study with proper design and implementation
- ✓Statistical controls through covariates (prior math scores)
- ✓Students nested within teachers via Multilevel Linear Model
- ✓At least one statistically significant, positive correlation with controls for selection bias
Conducted by The EdTech Collective & ICEI in partnership with Instructure