Learning Engineering for Online Education

Theoretical Contexts and Design-Based Examples

Chris Dede, John Richards, Bror Saxberg, Chris Dede, John Richards, Bror Saxberg

• Publisher: Routledge
• Year: 2018
• Print ISBN: 9780815394419
• eBook ISBN: 9781351186179
• Topic: Education
• Subtopic: Education General

Learning Engineering for Online Education is a comprehensive overview of the emerging field of learning engineering, a form of educational optimization driven by analytics, design-based research, and fast-paced, large-scale experimentation. Chapters written by instructional design and distance learning innovators explore the theoretical context of learning engineering and provide design-based examples from top educational institutions. Concluding with an agenda for future research, this volume is essential for those interested in using data and high-quality outcome evidence to improve student engagement, instructional efficacy, and results in online and blended settings.

1

Introduction

Improving Efficiency and Effectiveness Through Learning Engineering

Chris Dede

In the last two decades, digital media have greatly expanded the scope and impact of distance education, as well as widened the range of models used for teaching and learning. The rise of online learning as a major form of education for adults, coupled with the advent of MOOCs (massively open online courses), highlights the importance of enhancing digital education’s outcomes through improved instructional design. As a powerful method for accomplishing this goal, learning engineering applies a principled set of evidence-based strategies to the continual re-design of educational experiences to optimize their effectiveness and efficiency. This book centers on the use of learning engineering to improve online courses in higher education.

Both online teaching and blended instruction (online plus face-to-face) are good venues for learning engineering because the digital media used automatically generate rich, time-stamped log files documenting each student’s interactions with curricular materials, peers, and instructor. The evidence that guides constant, rapid cycles of improvement in learning engineering comes from this form of big data coupled with high quality outcome evidence (both near- and long-term). Doug Laney, an analyst with the META Group (now part of Gartner), described big data with a collection of “v” words (Laney, 2001), referring to (a) the increasing size of data (volume), (b) the increasing rate at which it is produced and analyzed (velocity), and (c) its increasing range of sources, formats, and representations (variety). To this, other authors have added veracity, to encompass the widely differing qualities of data sources, with significant differences in the coverage, accuracy, and timeliness of data (Dong & Srivasta, 2013).

Technological and methodological advances have enabled an unprecedented capability for decision making based on big data, and its use has become well established in business, entertainment, science, technology, and engineering (Dede, 2015). For example, online purchases are now guided by recommendation engines that analyze an individual’s shopping patterns and suggest products bought by others who have similar patterns of purchases. Big data is beginning to be utilized for decision making in higher education as well; one example is early identification of at-risk students based on analysis of their behavioral patterns. Thus far, these analytics focus on students’ macro-behaviors (e.g., adding or dropping courses) rather than their micro-behaviors (second-by-second activities in learning experiences coupled with evidence about learning outcomes).

Practical applications to analyzing students’ learning behaviors and outcomes in college and university instruction remain rare because of challenges unique to higher education (Dede, Ho, & Mitros, 2016). First, the sector lacks much of the computational infrastructure, tools, and human capacity required for effective collection, cleaning, analysis, and distribution of large datasets. Second, in collecting and analyzing student data, colleges and universities face privacy, safety, and security challenges not found in many scientific disciplines. Third, higher education should also be concerned with long-term goals—such as employability, critical thinking, and a healthy civic life—even though at many institutions these objectives are not apparent in their tactical decision making. Since it is difficult to measure these outcomes, particularly in short-term studies, researchers studying effectiveness in higher education often rely on theoretical and substantive arguments to justify imperfect, immediate proxies for these longitudinal objectives. This is made more difficult by a lack of training for or awareness among most faculty of the challenge of gathering valid and reliable evidence about learning outcomes, either near- or long-term.

Learning engineering cannot resolve these difficulties in measuring long- or short-term goals, but it can use big data to iteratively improve the design of learning experiences. Learning analytics and educational data mining are concerned with exploring the unique types of data that come from educational settings. Learning engineering combines methods from these fields with design-based research to better understand how students learn, what instructional strategies enable optimal learning (Baker & Yacef, 2009), and how to gather valid, reliable evidence about learners’ mastery of intended outcomes.

Learning engineering can improve online learning outcomes in higher education in a variety of ways. Through educational optimization, students’ engagement with courses can deepen, teachers can improve the efficiency of their instruction, and a broader range of learners can succeed because courses are tailored to their individual needs. The chapters in this book describe these and other types of improvements from learning engineering. Personalized learning is a conceptual framework that articulates the mechanisms that create many of these improvements.

Personalized Learning

The US Department of Education’s 2010 National Education Technology Plan provided an early, influential definition of personalization, stating:

Personalization refers to instruction that is paced to learning needs, tailored to learning preferences, and tailored to the specific interests of different learners. In an environment that is fully personalized, the learning objectives and content as well as the method and pace may all vary.

(US Department of Education, 2010, p. 12)

In the same time frame, the 2010 SIIA Symposium on [Re]Design for Personalized Learning articulated five essential elements of personalized learning: flexible, anytime/everywhere learning; redefine teacher role and expand “teacher”; project-based, authentic learning; student driven learning path; and mastery/competency-based progression/pace.

Since then, personalized learning has been clouded by many definitions, most of which weaken the concept to the point that everyone can claim they are already “personalizing” learning. (This bastardization of a rigorous educational innovation is unfortunately quite common.) For the purposes of contextualizing learning engineering, we define personalized learning as having four fundamental attributes:

1. Developing multimodal experiences and a differentiated curriculum based on universal design for learning principles;
2. Enabling each student’s agency in orchestrating the emphasis and process of his or her learning, in concert with the evidence about how learning works best and with mentoring about working toward long-term goals;
3. Providing community and collaboration to aid students in learning by fostering engagement, a growth mindset, self-efficacy, and academic tenacity; and
4. Guiding each student’s path through the curriculum based on diagnostic assessments embedded in each educational experience that are formative for further learning and instruction.

Substantial evidence exists that combining these four attributes leads to learning experiences that provide strong motivation and good educational outcomes for a broad spectrum of students (Dede & Richards, 2012).

As targets of opportunity for learning engineering in higher education, the strategies below for instructional interventions based on big data potentially provide several ways to improve learning through personalization (Dede, Ho, & Mitros, 2016):

• Individualizing a student’s path to content mastery, through adaptive, competency-based education. One example is using game-based environments for learning and assessment, where learning is situated in complex information and decision-making situations that adapt to each learner’s progress.
• Improving learning as a result of faster and more in-depth diagnosis of learning needs or course trouble spots, including assessment of skills such as systems thinking, collaboration, and problem solving in the context of deep, authentic subject-area knowledge assessments.
• Increasing the efficiency of learning to reduce overall costs to students and institutions.

The chapters in this book provide examples of applying learning engineering to these and other targets of opportunity.

The Value of Personalization in Online Learning

Throughout its history, distance education has struggled with efficiency and effectiveness due to a lack of personalization. A brief history of distance education illustrates this problem (Dede, Brown-L’Bahy, Ketelhut, & Whitehouse, 2004). In the 19th century, distance education in the United States was shaped by new technologies that allowed educators to overcome barriers of distance and time—shifting understandings of the purpose of education—and social, political, and geographic forces. The development and implementation of the first correspondence courses were credited to Sir Isaac Pitman of England, the inventor of shorthand. In 1840, he used the postal service in England to reach learners at a distance. A more formal version of the early American correspondence course was created by Anna Ticknor of Boston in 1873. In order to increase educational opportunities for women, she originated the Society to Encourage Studies at Home. The society provided courses of study for women of all social classes and served over 10,000 women over its 24-year lifespan (Nasseh, 1997; Stevens-Long & Crowell, 2002).

In 1878, John H. Vincent, co-founder of the Chautauqua Movement, created the Chautauqua Literary and Scientific Circle. This Circle offered a four-year correspondence course of readings; students who successfully completed the course were awarded a diploma. This course was open to all adults, including women and senior citizens (Scott, 1999). By 1892, the 19th century version of the “Information Superhighway” (otherwise known as rural free delivery) paved the way for Penn State University to provide higher education to rural families (Banas & Emory, 1998). Other institutions of higher education, notably the University of Chicago and the University of Wisconsin, modeled their extension schools after the Chautauqua program (Scott, 1999).

At the start of the 20th century, distance education still relied on correspondence courses delivered primarily through the postal service. Many of these courses were delivered to their students by mail, as discussed earlier, but did not allow much interaction or individualization (Moran, 1993). Although rules for home study were established in 1926 to allow some form of governmental control, correspondence methods were not conducive to supporting learners nor were they standardized (PBS, 2002). One of the main goals of early distance education was to help inculcate immigrants into the “American way of life” (Sumner, 2000), but these learners needed substantial guidance and aid. As a result, poor curricular design and lack of support were particularly problematic, and the dropout rate was high (Shea & Boser, 2001).

While distance education is rooted in the 19th century, the field blossomed in the 20th century. Distance educators looked to technological innovations to provide new opportunities for their field, and the 20th century was rife with technological advances (Mood, 1995). During the 20th century, distance education embraced radio, television, computers, and ultimately the Internet. As the methods of delivery for distance education expanded, so did the diversity of learners seeking distance education and their reasons for enrolling in such courses. Individuals interested in learning cultural norms, becoming more capable in the workforce, or hoping to re-situate themselves in their social context after wartime service, became major consumers of distance education (Dymock, 1995).

In the 1920s, distance education started to utilize radio for delivery of lessons (Bourke Distance Education Centre, 2002; Nasseh, 1997). In a push to widen access, speed the interaction between student and professor, and personalize the delivery of distance education, the use of radio was seen as an exciting opportunity. In the mid-1930s, an American art history course was offered by radio broadcasts (Funk, 1998), and other courses supported forming listening groups to enhance learning (Mood, 1995).

However, despite the rapid rise of radio technology, distance education courses were rarely if ever offered for credit in higher education (Nasseh, 1997). The education community, along with society as a whole, regarded legitimate education as only possible in conventional locales, such as classrooms (Funk, 1998). To address concerns about a lack of teacher interaction in distance education, a modification of the correspondence course was designed in Soviet Russia in the 1930s, called the Consultation Model (Tait, 1994). As its name implies, this type of correspondence course included periodic face-to-face meetings with instructors; however, unlike its name, the consultations were mostly lecture-based meetings intended to spread communist dogma.

Television was the next big advance in distance education technology. As early as 1934, the State University of Iowa used television to deliver course content. Early research into learning via television indicated mixed results, with several studies showing that it was similar to conventional instruction. Gayle Childs referred to televized distance education as an “‘instrument’ of delivery, not a pedagogical method” (Jeffries, 2002, p. 6).

Prior to the introduction of computer technologies in the 1960s, correspondence-course and independent-study models of distance education posed challenges to the learning and teaching processes. This contributed to a persistent problem of credibility for the field. Tele-courses (Verduin & Clark, 1991), which developed in the 1970s, showed promise for minimizing some of these problems. Previously, television had primarily been used as an electronic blackboard and for the delivery of standardized content through lectures intended to reach wide audiences. The development of videotape allowed educators to customize the same content for different learning environments. This medium also allowed increased flexibility; course content could be stored, delivered, and repeated at will. This minimized time-dependency, a drawback of previous televized courses. However, despite their advantages, the cost and complexity of producing tele-courses made them impractical for teaching large numbers of students.

Around the same time, the open university concept was launched. The creation of universities open to all was driven by the need to provide alternative education for adults whose needs could not be met in the traditional classroom. The British Open University began in 1969 through video broadcasting of its weekly courses on the BBC. Over time and with the advent of new technologies, the British Open University’s model of distance learning evolved into a student-centered delivery system and administrative structure separate from a campus setting. More economically practical than tele-courses, this system envisioned each student as “a node in the network” (Granger, 1990, p. 189) that provided individualized instruction in a virtual classroom. The students had access to a virtual library—customizable based on their particular learning style—and to collaborative tools that encourage discourse and critical thinking (Prewitt, 1998). By encouraging a community of learners, this model overcame some of the problem of isolation.

During the 1970s, the capability of computers to automate tasks and deliver information made them invaluable tools for many companies, thereby increasing the need for technologically competent workers. This prompted the inceptio...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Table of Contents
• List of figures
• List of tables
• List of contributors
• Foreword by Huntington Lambert
• 1. Introduction: Improving Efficiency and Effectiveness Through Learning Engineering
• PART I: Dimensions of Learning Engineering
• PART II: Cases of Practice
• PART III: Conclusions
• Index