Data

With the recent success of our pilot program, we saw some really exciting data. However, in truth our sample size was small and our experiment lacked scientific rigor, so we’d like to highlight some third party data that validates a lot of the theory behind Boost. Some of the most convincing studies we’ve come across are two papers by the Harvard Education Laboratory’s Roland Fryer and Will Dobbie, a pair of economists who’ve done extensive research on what works and what doesn’t work in education.

In their 2012 paper, the pair investigated successful strategies employed by charter schools in New York City, focusing on five: frequent teacher feedback, use of data to guide instruction, increased instructional time, high expectations and finally high-dosage tutoring. The authors noticed a strong correlation between the presence/robustness of tutoring programs and successful schools within their sample. Specifically, a school with tutoring had ELA (English Language Arts – a common core standards system) scores that were on average .04 standard deviations higher than a comparable school without tutoring. To determine the robustness of this effect, the authors ran a second study and published a working paper in 2013.

In this study, the authors applied these five strategies to underperforming public schools in Houston and examined the corresponding changes in state math and ELA scores. Interestingly, the authors noticed an even more pronounced effect. In the authors’ words, “students who received tutoring in secondary schools outperformed their peers by over 100 percent.” We at Boost are obviously pretty stoked about what this means for the potential effectiveness of the tutoring we can provide.

- Chris Hopkins