Randomized Controlled Trial (RCT) of...

Zearn Math, a Popular Software Platform, to Support Grade 3-5 Math Education

Reviewed

K-12 Education
Accurately reported

The Annenberg Institute (June 2025) posted an RCT of Zearn Math, a popular software platform, to support math learning for students in grades 3-5. This high-quality RCT found suggestive (not statistically significant) modest effects on state test scores in math after two years of program implementation.

Zearn Math is a popular software platform for K-8 mathematics learning, designed to enable all students to successfully access grade-level content. RAND researchers collaborated with Zearn, the product’s developer, to design this evaluation. Then RAND conducted the study independently, randomly assigning 64 schools in an urban Texas district to either supplement classroom instruction with Zearn Math in grades 3-5 for two years – or to continue with business-as-usual, which included various other supplemental technology products. High proportions of economically disadvantaged, Hispanic, English-learner, and below-proficient students made up the primary sample of 10,000+ students.  The study preregistered two confirmatory research questions about Zearn Math’s effects on Texas STAAR math assessment scores, for all students and students below proficient at baseline. Those results were positive but not statistically significant; equivalent to raising a control group student from the median to the 53rd or 54th percentile.  Although this study did not yield confirmatory evidence that Zearn Math improves student learning, consistent positive signals across all estimated confirmatory and exploratory effects, including on the MAP adaptive mathematics assessment, suggest it holds promise to do so.

Full Study Report

We have no suggested revisions to the study abstract.

No-Spin’s Study Overview

Large, high-quality RCT of Zearn Math – a popular software platform – to support math learning in grades 3-5 finds suggestive (not statistically significant) modest impacts on state test scores in math after two years of program implementation.

Program:​

  • Zearn Math is a software platform which, in this study, complemented existing classroom instruction in grades 3-5. Zearn provides students with in-class digital lessons that are designed to "synchronize with the grade-level material currently being covered during teacher-led instruction, and when students struggle, to provide individualized, targeted remediation to scaffold learning of the grade-level material."

Study Design:

  • The study randomized 64 schools in one large, urban Texas school district to either implement Zearn Math starting in grades 3 and 4 (treatment group), or continue math instruction as usual (control group).
  • Treatment group students received Zearn for two school years. Their use of Zearn was well below recommended levels in Year 1 but much higher in Year 2. In year 2, the median student completed the recommended 90 lessons and used Zearn 102 minutes per week (above the recommended 90 minutes). 
  • Of the 10,577 3rd and 4th grade students in the schools at study entry, 87% were Hispanic or Black, 67% were below grade-level proficiency in math, 83% were economically disadvantaged, and 45% were English language learners.
  • Based on careful review, this was a high-quality RCT (e.g. baseline balance, preregistered analyses, low sample attrition).

Findings:

  • ​Two school years after program entry –  when students were completing 4th and 5th grades –  the study found modest, but not statistically significant, effects on the STAAR state exam for both the full sample (effect size 0.07, p=0.22) and the subgroup of students who were below grade-level proficiency at baseline (effect size 0.10, p=0.20).   
  • The effect for the full sample represents approximately a 13% improvement over the annual gain in math otherwise expected for 4th and 5th graders.1 The effect for students with below grade-level proficiency represents approximately a 19% improvement over the expected annual gain.

  • Because these findings were not statistically significant, they constitute suggestive but not strong evidence of an effect.

1 The average annual gain in math achievement for U.S. fourth and fifth grade students on six nationally normed tests is approximately 0.54 standard deviations (see Hill, Bloom, Black, and Lipsey, 2007). 

Click or tap a highlight to see No-Spin’s comment