世界银行-数学补救的数字个性化学习能否为高等教育提供公平的竞争环境？厄瓜多尔的实验证据（英）-2023.6-30页_372kb

Title and Main Question: Can digital personalized learning (DPL) using software like ALEKS level the playing field in higher education by addressing math remediation gaps? The study investigates this through experimental evidence from Ecuador.

Context and Problem: Ecuadorian students entering higher education often lack foundational math skills, hindering academic success. Traditional remedial programs are scarce and costly due to institutional constraints. The COVID-19 pandemic exacerbated learning losses, making DPL a viable, cost-effective alternative for large-scale implementation.

Experimental Design: A randomized controlled trial was conducted in 39 technical and technological institutes, with 32 as controls. The ALEKS software was offered for six months to assess impacts on academic outcomes, including course repetition, enrollment, and math test scores.

Key Findings:

• Students using ALEKS showed a large, marginally significant decline in course repetition probability and a very large positive impact on math standardized test scores.
• No effect was observed on enrollment in the third semester.
• Heterogeneous effects: Male students saw larger reductions in repetition, likely due to higher enrollment in STEM fields, while female impacts were neutral. Field of study also influenced results, with substantial effects in engineering.
• Adoption and usage: About 84% of eligible students used the software initially, but engagement dropped over time, averaging around 90 minutes per week for many.
• Mechanisms: ALEKS increased study hours by approximately 10-15% and boosted perceived math ability without reducing time on other subjects.

Cost and Scalability: The intervention cost $18 per student, making it cost-effective. Potential for large-scale application exists, though outcomes may vary in different instructional settings, especially with a return to in-person learning.

Conclusion: DPL software like ALEKS can significantly improve math readiness and reduce course failure rates, offering a practical solution for higher education in resource-constrained environments. However, effects may differ by student demographics and educational contexts.