In a separate study, researchers Sarah Souders and Robert Bifulco examined a similar policy in Ohio. There, students took a reading test in the fall of third grade. Again, the study tracked students who just failed, and thus risked being held back, versus those who just passed. Over the course of that year, the at-risk students’ reading scores rose faster. Some of those gains even persisted into future grades, suggesting they reflected real improvements in learning, not exam cramming.