Educating More Students At-Scale: High Quality, Inclusive, and Online
13 Tips You Can Use Now
By Tania P. Heap and Adam D. Fein
Across the globe, students in the spring of 2020 had to adapt to non-typical ways of learning. Traditional Gen Z college students, full time-working professionals, single parents, students with disabilities, and international students all experienced unprecedented ways of living and studying during quarantine. Incoming freshmen want to continue their education but are interested in more affordable options. The challenges facing higher education before the global pandemic — particularly in public higher ed — haven’t dissipated, rather, have been exacerbated, as continued disinvestment from states, the looming 2026 enrollment cliff, and heavy dependence on location-specific income (dorms, food, Starbucks, recreational center, etc.) paint a gloomy picture for the near term. Yet there are changes that can be made, quality-centered adaptations, that proactive institutions should begin to consider immediately.
Did you know that 20 million (or two-thirds) of 18–24 year-olds are not enrolled in college? Over and above providing more options for the students we do educate, as we welcome new students to our institutions and consider that the preponderance of these students need to work and have family commitments, we will need to provide more and better flexible learning options. These options may need to be offered either fully or partially online and, at-scale — allowing the institution to lower the cost for students by increasing the number of students in a program. Lowering the cost of education is not something often considered when dealing with the challenges presented above. But high enrollment does not have to equal low quality. Our UNT-NetDragon Digital Research Center is hard at work on strategies for success when offering online courses at-scale.
Scalable online education is adaptable. Content, interaction, and assessment can be adjusted to fit into different modes of delivery and schedules (e.g., blended and online courses, low and high enrollment courses, short format offerings, courses targeting a local vs. global audience). These strategies, recommended for instructors and designers and outlined further in this article, enable content to be scalable, to adapt more easily to a range of learning environments and, most importantly, emphasize quality.
Yet, despite the fact that high-enrollment online courses and programs help reduce costs, concerns have been raised over student retention in both massive open online courses (MOOCs) and large enrollment credit-bearing classes. A combination of pedagogical strategies and artificial intelligence (AI) driven tools can be used to mitigate concerns over quality of learning experience. For instance, well-crafted grading rubrics can reduce instructor and teaching assistant (TA) grading workload by giving students the opportunity to interact and assess the efforts of their peers. Findings from research examining the impact of peer assessments in large online classes highlight the importance of equipping students with tools and skills to grade each other reliably and fairly, by providing students with concise and unambiguous instructions, sharing the grading rubric in advance, and allowing students to find familiarity by practicing and repeating the process.
Students can also play an important role in quality at scale by maintaining interaction and engagement in large classes. Coursera, the world’s largest MOOC provider, has a community mentors program consisting of experienced MOOC learners volunteering their time to moderate discussion forums. Arizona State University adopted a scaling model for high-enrollment online courses that leverage TAs, team-teaching through multiple instructors sharing the same instructional load, student leads supporting smaller groups of students in the course, and machine learning techniques for grading and feedback.
AI-based technologies have established a place in the success of at-scale learning and its role will grow dramatically in the coming decade. AI can help maintain the perception of an instructor-student interaction, which is crucial for student participation and learning. AI-driven TAs can reduce instructor workload for responding and participating in discussion forums within large blended classes. AI-based implementations are popular in mathematics and foreign language learning and have been used successfully in more subjective disciplines such as programs for automated scoring of writing quality. AI technology is also used to address academic integrity and online exam proctoring. Not relying on human proctors means that the student can schedule an exam any time any day from anywhere in the world without advance scheduling.
Synchronous components of a class are harder to adapt to the challenges presented by learners in different time zones across the globe, or working full time jobs at all hours. Education designed to be scalable can meet these challenges through a combination of media to present the same content. Text-based, downloadable, offline, audio- and video-based elements, and chunked content that uses platform-agnostic instructions are ways to offer instruction that (1) suits the varying needs of a broad learner audience, and (2) makes the content easier to be redesigned and transferred from: one platform to another, one teaching schedule to another, and one course length to another.
It is important to note that for these quality-driven strategies to truly be successful for students and institutions, they must work for all populations. It is well noted that online learning, particularly at-scale, can be more difficult for underrepresented populations. The demographics of students across the country are evolving: students are more diverse, there are more first-generation college students, students with fewer financial means, and more students who enter college underprepared. Some research is starting to show that well-designed online material that focuses on engagement and feedback and is built with flexible asynchronous components can be beneficial to underrepresented populations, while poorly designed online courses can be more detrimental to these populations.
Finally, much of the research and data we’ve examined on scalable online education comes from graduate-level programs, whose learner demographics can differ from undergraduate college students. This year, the University of North Texas (UNT) partnered with Design@Coursera to bring the Bachelor of Applied Arts and Sciences (BAAS) degree completion program to an online global student audience beginning in fall 2020. UNT’s BAAS program will offer a pioneering opportunity to the faculty fellows in our UNT-NetDragon Digital Research Center to explore the experiences of students and instructors in this undergraduate program designed to scale. Our goal is to continuously improve our own teaching and learning practices and share empirically driven, digital-first strategies among other instructors, researchers, and professionals interested in quality scalable online education.
If you are interested in collaborating or discussing research on scalable online instruction, feel free to contact Dr. Tania Heap (tania.heap@unt.edu) or Dr. Cassie Hudson (cassie.hudson@unt.edu) at the University of North Texas for more information. We would love to hear from you!
Tips for quality at scale
[Tip 1] Create stackable, modularized content that students can follow at their own pace.
[Tip 2] Include learner diagnostic activities (e.g., ungraded quizzes for knowledge testing) so that students can self-assess their readiness to take a course.
[Tip 3] Present content via multiple mediums (text-based, media-based, offline downloadable, etc.)
[Tip 4] Devise alternative strategies to enhance student-student and instructor-student interaction while keeping instructor intervention and workload to manageable levels (e.g., asynchronous discussion forums, AI-driven interventions such as Packback, peer-grading).
[Tip 6] Work with AI-based cognitive tutors that provide “hints” while students work through problems on their own.
[Tip 7] Utilize machine learning techniques for grading and feedback.
[Tip 8] Create well-crafted grading rubrics to reduce instructor and teaching assistant (TA) grading workload.
[Tip 9] Provide students with concise and unambiguous instructions for peer assessment.
[Tip 10] Engage alumni and subject matter experts as digital community mentors in large discussions.
[Tip 11] Team-teach with multiple instructors sharing the same instructional load.
[Tip 12] Deploy student leads to support and manage smaller breakout groups
[Tip 13] Use AI to address academic integrity and online exam proctoring.
