Can Digital Personalized Learning for Mathematics Remediation Level the Playing Field inHigher Education ? Experimental Evidence from Ecuador

Many Ecuadorian students entering higher education have cognitive skills gaps in mathematicsthat undermine their ability to assimilate academic contents. This paper presents the results of a randomizedcontrolled trial assessing the effects on academic outcomes of a Digital Personalized Learning Software for mathematicsremediation (the ALEKS software) offered to first-year students entering technical and technological highereducation programs in Ecuador amid the COVID-19 pandemic.The possibility to use the software led to a large and marginally significant decline in the probability ofrepeating a course, as well as a very large positive impact on standardized test scores in math. The analysis finds noimpact on the probability of enrolling in the third semester. When disaggregating the impacts, the findings showthat the effects on repetition are particularly large for male students, possibly because of higher male enrollment inscience, technology, engineering, and mathematics disciplines. When assessing the potential mechanisms, thefindings show evidence that the software led to a net increase in hours dedicated to studying mathematics. Theresults suggest that Digital Personalized Learning Software can be a cost-effective solution for math remediation withpotential for large-scale application.