Designing Active & Adaptive Learning for Large Online and Blended Courses: What We Got Right, What We Learned…

Concurrent Session 6

Brief Abstract

Since 2011, our university designed and implemented adaptive AND active learning experiences for large enrollment blended and online courses. This past academic year, enrollment grew to 27,000 learners. Results indicate increased student success; however, the transformation encountered growing pains. We'll share important lessons learned from the Instructional Design perspective.

Presenters

Peter van Leusen is the Manager of Instructional Design with focus on scalable and adaptive projects for EdPlus, Strategic Design & Development at Arizona State University. Over the past 16 years, he has worked at multiple large research-focus universities in the US with the goals to foster student success and truly have an impact. His recent instructional design projects include collaborating on adaptive curriculum and courseware development, designing MOOCs in a foreign language, and spearheading innovative digital educational experiences for broad audiences. Before joining ASU, Peter worked as the Assistant Director in the Office of Instructional Consulting in the School of Education at Indiana University and previously taught middle and high school German. Peter holds a PhD in Instructional Systems Technology from Indiana University, Bloomington. His research interests include Faculty Development, Instructional Technology, Instructional Design, and Innovative Teaching & Learning in face-to-face, hybrid, and online formats.

Additional Authors

Laura DePue is an instructional designer at Arizona State University. She began her career by teaching Comparative Religious Studies at the college-level in both traditional and online formats. After continuing to improve her skills in the area of online teaching, she began working as an instructional designer full-time, and has specialized in the areas of accessibility, active learning, gamification, and adaptivity.

Extended Abstract

 

Adaptive learning offers promise for all learners of differing skills and abilities to be successful at their individual pace and level of understanding (ELI, 2017; Pugliese, 2016). As an institution centered on student success and access, our university started implementing adaptive courses in 2011 with enrollment in adaptive courses reaching 27,000 learners in 2018-2019.

 

Our university and the vendors with whom we work have been implementing active and adaptive learning together in general education courses at increasing scale. This session summarizes outcomes (student retention, success, and instructor and student feedback) from several subject areas (Economics, Psychology, US History, and Biology), and describes current efforts for developing an increasingly personalized and programmatic curriculum.

 

Despite the promise of student success and growth in vendors, adoption of this model is slow in higher education (Bryant, 2016). Common barriers are few implementations beyond pilots, technical challenges, and faculty skepticism. In contrast, our university approached its adaptive implementation from a systems view - where organizational systems are interrelated and changes impact other sub-systems (von Bertalanffy, 1956).

 

To start successfully, the initiative aligned with our university's overall mission of student success. Empirical evidence focuses on two statistics: (1) retention (start/completion of courses); and (2) success (grade C or better). For example, results from two semesters of two economics courses show that withdrawal rates decreased by 50% (Macroeconomics) and 33% (Microeconomics) and student success rates (A, B, C grades on exams) increased by 25% and 15% respectively. Furthermore, time spent in online learning environments increased by 300% in Macroeconomics compared to previous semesters. The improvement is attributed to a flipped classroom, active + adaptive model.

 

Secondly, to facilitate changes in academic practices, faculty development focused on data-driven decision making and transforming lecture-based instruction to active learning. To help visualize the implementation of the adaptive and active model, one might think of splitting a typical lesson into two parts: (1) the adaptive learning part, to be completed individually by students before class; and (2) the active learning part, student-centered techniques to be completed individually and in groups, either in the classroom or in the Learning Management System (LMS). The combination of the adaptive learning system with in-class active learning takes students through the full range of Bloom's Taxonomy learning levels, where lower levels (i.e., knowledge, comprehension, and application) objectives are primarily addressed in the adaptive learning system and the higher levels (i.e., analysis, evaluation, and creation) are primarily addressed in the classroom and/or LMS.

 

An adaptive learning courseware is an online platform that "dynamically adjusts [learning materials] to student interactions and performance levels, delivering the types of content in an appropriate sequence that individual learners need at specific points in time to make progress" (ELI, 2017, p. 1). Early attempts to provide personalized instruction by machines can be traced back to B.F. Skinner's teaching machine in 1954 (Skinner, 1958). According to Skinner (1958), teaching machines provide "optimal conditions for individual instruction" (p. 969). In addition to massive advances over the past decades, such as the creation of the Internet, recent technological developments in database management and data analytics offers the opportunity to better adjust assessments, feedback, content, and various media to student behavior and knowledge (ELI, 2017).

According to Johnson (2017), adaptive courseware today allow learners, who are enrolled in the same course,  to experience different content in variable sequences. In particular, these systems guide and recommend content and activities based on four types of adaptivity: (a) algorithm, (b) assessment, (c) association, and (d) learner agency. Algorithms, drawing from the power of learning analytics (big data), analyze the behavior and performance of previous users of the courseware. As patterns are established, algorithms can recommend helpful content items to current learners. Assessment, or rapid remediation, is the most simple form of adaptivity. As each learner interacts with specific material, she or he encounters an assessment activity. Based on the learner's performance, the courseware either shares a supplemental resource to help learners be successful in the next attempt, or provides access to new material. Association is available in courseware that have multiple predetermined learning sequence or paths established. At some point in the learning experience, users make a choice on what learning path to pursue. Finally, agency allows learners to  self-select which topic or activity to pursue at any time. Similar to an open book, learners can search through the content items and choose a topic of their interest.

Considering these four types of adaptivity, learners enrolled in a course that employs adaptive learning courseware could receive a different - a more personalized - experience relative to a course which relies on traditional instruction methods.

In addition to using the adaptive courseware, the previously lecture-based format was changed to incorporate active learning techniques aimed at fostering higher-order thinking skills associated with higher levels of learning found on Bloom's Taxonomy. Active learning techniques utilize instructional activities, "in which students engage the material they study through reading, writing, talking, listening, and reflecting" (University of Minnesota, 2015). These new structures permit students to achieve higher order skills and reach deeper learning effects while honing students’ ”soft” skills, valued by today’s dynamic and evolving job market. The new lesson design, or “Applied Concept Exercise” (ACE), guides students through five learning stages during each class. The learning stages Engage, Explore, Explain, Elaborate, and Evaluate are based on the 5E Instructional Design model (BSCS, 2006).

Since previous research indicated that incorporating active learning into a class requires more time than delivering material via lecture (Zappe, Leicht, Messner, Litzinger, & Lee, 2009), the faculty and Instructional Designers also decided to utilize a format where students engage individually with the adaptive platform before coming to class. Not only is this "flipped" model recommended by numerous research studies (e.g., Jensen, Kummer, & Godoy,  2015; Roach, 2014; Roehl, Reddy, & Shannon, 2013), it had also shown positive improvements regarding student performance and retention in previous implementations at our university.

Thirdly, to address technology barriers, the presenters established strategic partnerships with adaptive vendors. Through partnerships, faculty and designers are directly impacting the course design, technical capabilities, and reporting functions of these platforms. In addition, to facilitate active learning in large lecture halls, various technology solutions were explored and used (e.g., student response systems). All systems had to be integrated with the Learning Management system to efficiently teach at scale.

To conclude the session, the presenter will share proven implementation models for developing adaptive courseware and implementing active learning in a blended and online setting. Experience will be summarized from an Instructional Design perspective, along with lessons learned when collaborating with faculty, student success outcomes, and learner perspectives from multiple subjects.

 

Plan for Interactivity: Engaging audience through polling (Poll Everywhere or Slido)

 

To engage the audience in meaningful and effective ways - as well as to address the adaptive spirit of the topic - the session includes multiple opportunities to gauge interests, share experiences, and customize the content of the presentation as needed. In general, the engagement occurs in two formats: (1) in person; and (2) using a polling system. In person, the presenters will utilize traditional engagement techniques that require audience participation. Preference is given to activities that gauge the audience interests and allow for peer-to-peer communication, such as Think-Pair-Share.  Additionally, this session uses a polling software (such as Poll Everywhere or Slido) that is monitored by one of the presenters. Attendees can utilize this modality to provide comments or ask questions to presenters and peers. Furthermore, due to the large scope of this topic, the presenters provide opportunities at key points that allow the audience to adapt or customize the session to better meet their interests. For example, when discussing the implementation model of adaptive learning, the presenters poll the audience to find out which course or sub-topic is most of interest. Based on results, the presenters focus on the subject area that received most responses. Finally, to continue the discussion beyond the session, the presenters posts additional questions related to adaptive learning that require the participants to reflect on their own contexts.

 

Takeaways/objectives:

 

  1. Summarize outcomes, including successes and lessons learned, from implementing adaptive and active learning courses.

  2. Describe the implementation of adaptive and active learning in large blended and online courses.

  3. Illustrate a systems view for achieving student success.