Online Learning in Higher Education: Evidence, Growth, and Quality Questions

Online learning in higher education has grown from a marginal phenomenon in the 1990s to a mainstream delivery mode that serves millions of students. The COVID-19 pandemic forced the rapid shift of virtually all higher education online, creating both the largest natural experiment in online learning and a period of crisis that produced very different outcomes from well-designed online courses. Understanding the evidence on online learning effectiveness, which conditions support it, and what institutions have learned from rapid growth and forced adoption, is essential for higher education policy and institutional strategy. The evidence on online versus in-person learning outcomes is complex and context-dependent. Meta-analyses of research conducted before the pandemic generally found that online learning produced outcomes comparable to or slightly better than in-person learning, but with very high variance across studies. The variance reflects that online learning is not a single intervention but a wide range of delivery models, instructional designs, and student support systems that differ dramatically in their characteristics. A well-designed online course with active learning elements, regular instructor interaction, and adequate student support is fundamentally different from a video-lecture-based course with minimal student-instructor interaction. Student characteristics significantly moderate online learning outcomes. Research consistently finds that online learning is associated with lower completion rates than in-person learning in community college settings, particularly for students who are less academically prepared, working more hours, or lacking the self-regulation skills that online learning requires. These populations, which include many of the students who stand to benefit most from expanded access, are the ones for whom online learning is most challenging without adequate support. Synchronous and asynchronous online learning differ in their interaction dynamics. Synchronous learning, which occurs in real time through video conferencing, preserves some of the interaction dynamics of in-person classes but requires scheduling coordination that may disadvantage students with work or family obligations. Asynchronous learning, which allows students to engage with course content on their own schedule, provides maximum flexibility but requires stronger self-regulation skills and more deliberate design of student engagement. Research suggests that well-designed asynchronous courses can produce comparable outcomes to synchronous courses, but that both require thoughtful design to support engagement. Hybrid or hyflex models, which combine elements of in-person and online delivery, have grown in interest as institutions seek to offer flexibility while maintaining in-person community. These models are logistically complex and pedagogically demanding, requiring instructors to design for two modes simultaneously. Research on hybrid learning outcomes is more limited than for fully online or fully in-person, and implementation quality varies enormously. Completion rates in online courses and programs have been lower than in in-person equivalents, a disparity that is more pronounced in some fields and student populations than others. Research identifies several factors associated with lower online completion: lack of social connection, reduced accountability that comes with not being physically present, technical difficulties, and insufficient access to student support services online. Institutions that have invested in online student support services including virtual advising, online tutoring, and proactive outreach to at-risk students show better completion outcomes than those that treat online students as self-sufficient. Massive open online courses, or MOOCs, generated enormous initial enthusiasm about democratizing access to high-quality education from elite institutions. Research on MOOC completion and learning outcomes has consistently shown very low completion rates, typically 5 to 15 percent of enrolled learners, though the appropriate denominator for completion rate calculations has been debated given the diversity of learner intent. MOOCs have not delivered on their most transformative access promises, partly because their pedagogical model was not designed for the less-prepared learners who would benefit most from expanded access. Quality assurance in online learning has been addressed through accreditation standards, institutional policies, and instructional design guidelines. The Quality Matters rubric, developed specifically for online courses, provides a research-based framework for evaluating online course design against established standards. Institutions that have adopted structured quality assurance processes for online courses show better student outcomes than those that have not. The question of equity in online learning deserves ongoing attention. The digital divide, which affects access to broadband, devices, and digital skills, creates barriers to online learning that are not uniformly distributed. First-generation students, students from lower-income backgrounds, and older students may face particular challenges with online learning environments that assume technological fluency and self-directed learning skills. Institutions that monitor equity in online learning outcomes and invest in closing documented gaps produce more equitable results than those that assume equal access to the online environment.