We’re excited to share the results of the first quasi-experimental study (ESSA Tier II) on Once.
A quasi-experimental study estimates the causal impact of an intervention on a target population without using random assignment, but by matching treatment students to control students with similar demographics and similar BOY proficiency.
This study was conducted by LXD Research in a large southeastern school district on data from the 2024-25 school year. Fastbridge, by Renaissance Learning, is the nationally normed, computer-adaptive assessment that was used to measure growth.
The study had several significant findings. We’re most excited about the finding for kindergarten students who started the school year below the 50th percentile nationally on Fastbridge and received 80 or more Once tutoring sessions. This, we believe, measures the effect of the treatment on the treated, since prior correlational (ESSA Tier III) research has shown a disproportionate impact from Once on students who receive over 80 tutoring sessions compared to students who receive fewer than 80 tutoring sessions.
The effect that was observed here was 0.46 standard deviations with a p-value <0.001, meaning there is a less than 0.1% chance that the effect was the result of luck; the finding is statistically significant. The effect size is substantially larger than what other high-quality tutoring programs have reported in quasi-experimental studies and randomized controlled trials.
To be clear, this was a quasi-experimental study, not a randomized controlled trial, so the impact could have been confounded by other non-random factors that the researchers were not able to control for.
We have a separate randomized controlled trial in progress during school year 2025-26, and we look forward to sharing those results in the next several months.
The report linked below is a pre-publication copy that may contain typos that will be corrected before publication.
Once — Quasi-Experimental Study Fastbridge Report 24-25, LXD Research (PDF)