NCAT Monograph

Innovations in Online Learning: Moving Beyond No Significant Difference

By Carol A. Twigg

Download PDF version

I. Individualization: The Key to Innovation
II. Improving the Quality of Student Learning
III. Increasing Access to Higher Education
IV. Reducing the Costs of Teaching and Learning
V. Sustaining Innovation
Case: University of Illinois at Urbana-Champaign, Master of Science Degree: LEEP3
Case: University of Central Florida, Reactive Behavior Patterns: Implications for Web-based Teaching & Learning
Case: Rio Salado College, A Systems Approach to Online Learning
Case: Cardean University, Problem-centered Pedagogy
Case: Ohio State University, A Buffet of Learning Opportunities
Case: University of Phoenix, A Focus on the Customer
Case: Rio Salado College, Online Human Anatomy
Case: Excelsior College, What You Know Is More Important than Where or How You Learned It
Case: Drexel University, Modularizing Computer Programming
Case: The British Open University Approach to Online Learning
Case: University of Illinois at Urbana-Champaign, The Spanish Project
Case: Virginia Polytechnic Institute and State University, The Math Emporium: Student-paced Mathematics 24x7
Case: Michigan State University, CAPA: Computer-assisted Personalized Assignment System
Symposium Participants


During the early 1990s, many of those interested in the impact of information technology liked to talk about "paradigm shifts." Despite its attainment of cliché status, the concept of a paradigm shift is a powerful one. Most who were once skeptical of the impact of the Internet on the ways we do business in all facets of society now recognize that our paradigms are, in fact, shifting.

The word paradigm comes from the Greek word paradeigma, meaning "model" or "pattern." A paradigm represents a way of looking at the world, a shared set of assumptions that enable us to understand or predict behavior. Paradigms have a powerful influence on individuals and on society because our view of the world is determined by our set of assumptions about it. To put it another way, our vision is often affected by what we believe about the world; our beliefs often determine the information that we "see."

Extending this concept to technology, a paradigm effect may prevent people from seeing what is happening around them and from realizing the potential in a new application of technology. As Jim Wetherbe, Bobby G. Stevenson Chair in Information Technology at Texas Tech, puts it, "The biggest obstacle to innovation is thinking it can be done the old way." Familiar examples of how, in Wetherbe’s words, "technique lags behind technology" come to mind:

  • Faced with the invention of the telegraph, the Pony Express initially responded by buying faster horses. When that failed, the organization tried to hire better riders. It did not realize that the world had changed, and the Pony Express went out of business.
  • Shot from a single fixed position while actors paraded in front of the camera, early motion pictures were essentially stage plays on film. In 1903 The Great Train Robbery introduced narrative storytelling to films along with parallel action. Filmmakers intercut two or more stories taking place at the same time shot from different camera positions and distances, and an entirely new art form was born.
  • The first ATM was located inside a bank and was available only during banking hours. Bankers viewed this technological innovation as an automated teller. Real innovation did not occur until ATMs were placed outside banks and in malls, grocery stores, and airports, available twenty-four hours a day.

As we enter the new millennium, colleges and universities are offering thousands of online courses and, in the process, are ostensibly altering centuries-old methods of teaching and learning. Some would argue that this represents a paradigm shift. But does it?

There is no question that the higher education community has moved well beyond the time-and-place-specific campus paradigm of the 1980s and early 1990s, when discussions of IT applications consisted primarily of wiring the classroom or wiring the campus. Most of those engaged in online learning programs promote the benefits of 24/7 access to courses and degree programs. Because they may not need to go to campus as frequently or at all, students also value the flexibility offered by online programs. A lot has changed.

At the same time, a lot has not changed. The vast majority of online courses are organized in much the same manner as are their campus counterparts: developed by individual faculty members, with some support from the IT staff, and offered within a semester or quarter framework. Most follow traditional academic practices ("Here’s the syllabus, go off and read or do research, come back and discuss."), and most are evaluated using traditional student-satisfaction methods. This is hardly surprising, since most online courses are offered by traditional institutions of higher education. To return to our paradigm discussion, a paradigm provides boundaries for behavior, guidelines for action, and rules for success. All paradigms give practitioners a worldview that enables them to solve specific problems. The higher education paradigm, honed and perfected for hundreds of years, has served us well.

Leaders of the old paradigm community have a tremendous amount of time and energy invested in using the old rules. Consequently, they are often resistant to change and less likely to look for creative, innovative approaches to new opportunities. In much the same way that Thomas Kuhn (who first called our attention to the idea of paradigm shifts) observed scientists trying to "save the theory," so too do defenders of the old paradigm focus their efforts on old solutions to new problems.

The problem with applying old solutions to new problems in the world of online learning is that these applications tend to produce results that are "as good as" what we have done before—what is often referred to as the "no significant difference" phenomenon. Thomas L. Russell’s compendium of more than 355 comparative research studies suggests that students in technology-based (typically, distance learning) courses learn as well as their on-campus, face-to-face counterparts ( These studies have typically been used by distance educators to defend the quality of their courses and programs against the once-predominant view that learning takes place only in a physical classroom. What we need now, however, are new approaches that go beyond producing no significant difference.

On December 8–9, 2000, in Phoenix, Arizona, we gathered a group of faculty and administrators—those who were already "moving the ATMs outside the bank," so to speak—to consider the question of how to move online learning beyond being "as good as" traditional education. Before our meeting in Phoenix, we asked participants to think about how information technology can be used specifically to address the major challenges of higher education: improving quality, increasing access, and reducing costs. This paper, like the symposium discussion, is organized as a response to those objectives.

As we began our discussion in Phoenix, we asked symposium participants to do three things. The first was to analyze their assumptions about distributed learning. For example, although it is generally acknowledged that the more-effective online learning environments are learner-centered, there is much controversy and disagreement about what "learner-centered" means.

Advocates of "community" may demand residencies or synchronous online sessions, sincerely believing that such activities are learner-centered. Others view asynchronous learning environments as a keystone of learner-centeredness because such environments offer students greater flexibility. Is asynchronous communication de rigueur if one is learner-centered, or is synchronous exchange an important part of the learning experience? All too frequently, even innovative institutions fall back on a one-size-fits-all approach ("All of our student must do . . ."), forgetting that students are different and have different needs. What do we really mean by being learner-centered?

The second thing we asked the symposium participants to do was to step out of their paradigms and identify the strengths of each of the distributed learning approaches that we discussed at the symposium—especially in regard to particular kinds of students or particular academic topics—rather than advocating for one approach versus another. Are there some general principles that distinguish more innovative approaches?

Third, we asked the participants to explore what needs to be done to improve online education. Rather than comparing online learning with traditional higher education, how can we identify new models and talk about what is better rather than what is "as good as"? What are the important variables that create a rich online learning experience, one that makes real improvements in academic practice? How can each of us learn from others’ approaches and borrow aspects that can be integrated into our own learning environments?

A few words about terminology are in order. Throughout this paper, the terms distance learning, distance education, distributed learning, and online learning are used more or less interchangeably. At times, the use of distance learning seems appropriate because the issues under discussion most frequently concern off-campus (distance) versus on-campus learning. At other times, particularly when describing the new higher education environment, the phrase distributed learning more clearly expresses the changing nature (and the blending) of all forms of higher education. In any event, the reader should not draw unwarranted conclusions from a particular usage.

There is a saying among aficionados of Thoroughbred racing: "It’s not how fast you run; it’s how you run fast." Our goal in this paper is to show that it’s not providing student services online; it’s how you provide student services online. It’s the difference between online office hours and Rio Salado College’s "Beep a Tutor" idea: immediate on-demand help for students having learning problems. It’s the difference between a campus bookstore that mails books to distance learners and a professor who provides a one-click link on a course Web site to a particular page so that students can order the required book.

As you read this paper, we urge you to ask yourselves whether you are taking advantage of the capabilities of information technology in general and the Internet in particular as you design online learning environments or whether you are simply migrating your on-ground approaches online. Only by doing the former will we move beyond "no significant difference."

I. Individualization: The Key to Innovation

One can think of distributed learning programs as existing on a continuum from rather traditional, teacher-led distance learning programs on the one end (e.g., faculty teaching via television, faculty putting their courses on the Web, faculty leading computer-conference-based seminars) to more innovative, learner-centered programs that rely on a combination of high-quality, interactive learningware, asynchronous and synchronous conversations, and individualized mentoring on the other end. The former programs follow traditional schedules and structures (e.g., semesters, group meetings), may be delivered to fixed sites or involve residency requirements, and tend to be developed primarily by individual faculty members with appropriate IT support. The latter are modularized and self-paced, may include group experiences as appropriate and desirable, are delivered anywhere (sites, homes, and workplaces), diagnose students’ skill and knowledge level as they begin their programs of study, award credit for learning acquired outside formal educational structures to enable students to move more quickly through their programs, and are developed by teams of faculty, instructional designers, learning theorists, and IT staff, sometimes in partnership with commercial providers.

An example of a well-regarded traditional online program is the Master of Science degree in Library and Information Science (LEEP3) at the University of Illinois at Urbana-Champaign (see University of Illinois at Urbana-Champaign. Master of Science Degree: LEEP3). Courses and programs on the teacher-led end of the spectrum emulate face-to-face pedagogies and organizational frameworks, striving to make their quality equivalent to that of on-campus offerings. Programs such as LEEP3 do a fine job of replicating high-quality campus experiences. But do they go as far as they might in making significant improvements in the cost, the quality, or the access dimensions of student learning? Do they take full advantage of the inherent strengths of the Internet, which enables greater flexibility, convenience, and personalization?

A fundamental premise of this paper is that as long as we continue to replicate traditional approaches online—and continue to treat all students as if they were the same—we will once again find the "no significant difference" phenomenon vis-à-vis quality, and we will make only a negligible dent in the access problem rather than taking full advantage of the networked environment. And because these approaches bolt on technology to traditional teaching approaches, they will fail to reduce costs and, indeed, will frequently increase overall cost.

Despite the fact that the higher education community tends to treat quality, access, and cost as three separate and distinct issues, they are very much intertwined. It is now widely recognized, for example, that higher education’s historical approach to increasing quality—adding more faculty, more facilities, more resources—has simultaneously increased costs. We also know that access will be directly affected if the cost of higher education to students continues to rise. So too does a one-size-fits-all definition of academic quality limit access for students who bring diverse preparation, abilities, and interests to each learning experience. Conversely, because these three issues are so inextricably linked, there may be ways to address all of them simultaneously by using information technology. This paper gives example after example of how a particular approach to improving quality can also reduce costs while increasing access. Even though the issues of quality, access, and cost are addressed in separate sections below, the interrelationships among them will become apparent.

As we think about how to design more effective online learning environments, one thing is clear. We need to treat students as individuals rather than as homogenous groups. Rather than maintaining a fixed view of what all students want or what all students need, we need to be flexible and create environments that enable greater choice for students. Participants know from their own experience that students differ, for example, in the amount of interaction that they require with faculty, staff, or one another. At the British Open University, for example, approximately one-third of the students never interact with other people but pursue their studies independently. New York’s Excelsior College reports that 20 percent of its students take up to 80 percent of staff time, indicating a strong need for human interaction, in contrast to the 80 percent of students requiring very little interaction.

A number of institutions, like the University of Central Florida, are trying to understand possible relationships between students’ learning styles and online course development and delivery as well as the implications of that understanding for how we design online learning environments (see University of Central Florida. Reactive Behavior Patterns: Implications for Web-based Teaching & Learning). In a recent paper, UCF researchers summarized a number of studies that have examined the learning styles of students who enroll in distance education courses:

  • Boverie, Nagel, McGee, and Garcia (1997) incorporate the Kolb Learning Style Inventory (1998) into their study of learning styles, emotional intelligence, social presence and their relationship to satisfaction with distance education. They conclude that only social preference exists as a significant predictor of course satisfaction.
  • Tyler and Baylen (1998) use the Learning Styles Exercise developed by Kiersey and Bates (1978), finding the majority of their Web-based students are extroverted and judging, contrasting strongly with the instructor’s preference for introversion and perceiving. They speculate that differing perceptions of courses may be explained by contrasts (and potential conflicts) in learning styles between the instructor and students.
  • James and Gardner (1995) propose that learning styles are cast within a perceptual, cognitive, and affective framework, and suggest instructional design components for distance education that conform to learner needs within those three components.
  • Verduin and Clark (1991) argue that attention to the mode of learning preferred by students is important to the instructor who is designing distance learning experiences. They cite that Canfield (1983) developed a learning style model and instrument that bears relevance to online learning, and suggest that maturity has relevance in learning style considerations.
  • Ross and Schultz (1999) make recommendations for the interaction of online learning and learning styles relying on the theories of Dunn and Dunn (1978) and Gregoric (1982). They make specific suggestions for teaching and learning activities that conform to learning preferences of students.1

One implication of this research is that we need to think more creatively about how to develop course designs that respond to a greater variety of learning styles rather than concluding that online learning is more suitable for one type of student than another. The University of Central Florida has determined, for example, that the passive-independent Long type is more at risk in UCF’s online courses than are other types of students.
Because certain types of students respond more positively to today’s versions of online courses, some institutions have thought about counseling students who may not be successful not to take online courses. Instead, we need to be more thoughtful about course design so that we include structures and activities that work well with diverse types of students.

Taking this approach rather than limiting enrollment in online courses for some students requires real change, since it requires us both to understand our students as individuals and to offer many more learning options within each course. This paper is structured around a series of case studies presented by symposium participants. Some of these cases deal with courses, others with degree programs, and still others with institutions. At the symposium, participants described how their courses, degree programs, or institutions are trying to move beyond the "no significant difference" phenomenon by breaking away from the one-size-fits-all approach of traditional environments, whether on campus or online. We call these paradigm shifters the new providers.

All of the cases address increasing quality, improving access, and reducing costs to one degree or another, some more so than others. Each was selected because its approach to online learning is in some way differentiated from the instructor-led, semester-bound "traditional" approach that is predominant in higher education today. As a whole, they are characterized by such things as flexible enrollment options for students; personalized, on-demand, 24/7 student services; innovative curricular design that includes a focus on applied or problem-based learning taught by practicing professionals; and learner assessment that is integrated throughout the curriculum by diagnosing students’ knowledge and skill levels as they begin their programs of study and by responding accordingly.

Among the new providers, we distinguish between the groundbreakers, or those who have been leaders in breaking away from traditional approaches in many respects, and the new pacesetters, or those who have moved further along the continuum toward greater individualization for students.

No institution, program, or course described in the case studies has moved as fully along that continuum as is possible—and some have done more in one arena than another—but each illustrates a way to think about moving beyond the "no significant difference" phenomenon as we gain greater experience and knowledge about the intersection of online learning and the individual needs and interests of our students.

II. Improving the Quality of Student Learning

When asked about their views on the quality of online learning, most people in higher education begin by comparing what occurs in an online course with what goes on in the traditional classroom. A common assumption is that online learning cannot measure up to the in-class environment. In contrast, because of their years of direct experience with online learning, the symposium participants began their discussion about quality with the conviction that online learning is certainly as good as classroom learning. Rather than trying to compare one format with the other, symposium participants spent most of their time discussing the following question: What kinds of approaches to online learning will improve the quality of student learning? Consequently, they were able to come up with many ideas about how to improve quality by taking advantage of the capabilities of information technology and the Internet. In doing so, they considerably broadened what we mean by a "high-quality" learning experience. This new concept of quality takes us far beyond what is possible in a conventional classroom.

A fundamental premise of the symposium is that greater quality means greater individualization of learning experiences for students. This means moving away from teaching and learning ideas that begin with the thought that "all students need …" Information technology enables us to meet the needs of diverse students when, where, and how they want to learn. When we think about how to utilize technology to improve learning, the key is to focus on what we can do with IT that we cannot do without it. Technology can create environments that provide individualized learning approaches that serve each person in ways that he or she can most benefit.

Many of the leading institutions described in the cases in this paper tend to be attached to one way of doing things (e.g., synchronous versus asynchronous approaches). They thus illustrate pieces of the puzzle, if you will. Yet we are moving toward an online environment that radically increases the array of possibilities presented to each individual student. The ability to customize the learning environment so that each student can achieve in a variety of ways increases the likelihood that learning success online will be higher than learning success in the traditional classroom, dominated by a one-size-fits-all approach. Thus, the "right way" to design a high-quality online course depends entirely on the type of students involved.

Most of today’s online courses consist of putting the faculty member’s course online. These "traditional" online courses, much like their campus counterparts, are developed and delivered by individual faculty members, with some support from IT staff. Most follow traditional academic practices ("Here’s the syllabus, go off and read or do research, come back and discuss."), and most are evaluated using traditional student-satisfaction methods.

All of the new providers described below use technology to create a learning environment that is quite different from the traditional model. As one symposium participant put it: "We do not put the faculty member’s course online. Rather we use the faculty member’s expertise to define the learning outcomes, the applications of that learning, the content, and potential difficulties that students may encounter." Rather than trying to replicate a teaching model online, the idea is to create what has been called a "resource" model, an environment in which students interact and wrestle with learning materials directly (or in teams), under the tutorial guidance of a mentor.

Both the groundbreakers and the new pacesetters agree that students (either directly or in teams) need to interact with learning materials that allow them greater choices of assignments and resources. The key goal is for the students to become engaged in active "doing" in the learning process—that is, to move beyond merely reading text. Where the two kinds of new providers part company is the level of individualization to which the course aspires. While increasing the quality of courses from an instructional-design perspective, the groundbreakers tend to maintain a one-size-fits-all approach, holding to the conviction that their particular model is the "best." In contrast, the new pacesetters create a far richer learning environment in which students may make a variety of choices that meet their particular learning needs.

The Groundbreakers

Rio Salado College, the University of Phoenix, the British Open University, and Cardean University share a common approach to course development and delivery. Rio Salado’s "systems approach" typifies this model.

Individual courses at groundbreaking institutions are designed in the context of clear goals and desired learning outcomes set by content experts. The learning activities required of students are well thought out and correspond to what we know about human learning. As an example, UNext’s Cardean University business courses are designed based on the "learning by doing" philosophy of John Dewey and on current social constructivist views of learning.

The groundbreakers make several significant gains in quality when compared with those institutions using the traditional method of putting courses online. First, the level of the instructional design, including both pedagogical and technological aspects, is greatly increased. Rather than the single-source ("do your own thing") instructional development process employed by most institutions, the groundbreakers involve teams of experts in course development. Second, quality-control processes are more centralized, more collegial, and more elaborate than those in the traditional approach. Finally, course support structures, both during development and during delivery, are tightly integrated with the courses themselves, so that both students and faculty are assured of rapid responses to their needs.

The New Pacesetters

Virginia Tech, Drexel University, and Ohio State University, all part of the Pew Grant Program in Course Redesign,2 are developing new approaches that radically increase the quality of both the students’ learning experience and the learning outcomes achieved. Ohio State uses a buffet analogy to capture this new approach to online learning.

Many believe that mass customization is emerging as the organizing business principle of the twenty-first century. Internet-based e-commerce now makes it possible, for example, for customers to order computers designed to their exact needs and specifications, obtain customized home mortgages, and compile music CDs containing any combination of songs. By offering students a buffet of learning opportunities that can be customized to their learning needs, Ohio State, Virginia Tech, and Drexel University are pointing the way to a radically new approach to online learning.

Courses offered by the new pacesetters have five key features that can improve the quality of student learning:

  1. An initial assessment of each student’s knowledge/skill level and preferred learning style
  2. An array of high-quality, interactive learning materials and activities
  3. Individualized study plans
  4. Built-in, continuous assessment to provide instantaneous feedback
  5. Appropriate, varied kinds of human interaction when needed

1. Assessment of Knowledge/Skill Level and Learning Style
The first step in creating an individualized learning environment is to assess each student’s entering skill and knowledge level as well as his or her preferred learning style. Florida Gulf Coast University (FGCU) offers an introductory general-education course called "Styles and Ways of Learning." In that course, students complete the Myers-Briggs Type Indicator (MBTI) instrument, which identifies students’ preferences among sets of mental processes or mental habits. The MBTI makes students aware of the various ways in which they engage the world most successfully—for instance, through collaborative or individual experiences and through hands-on or intellectual processes. In its redesign of its introductory art-appreciation course, "Understanding the Visual and Performing Arts," FGCU will create learning activities that build on differences in students’ learning styles so that students can be directed to the learning activities most suited to their preferred learning styles, thus giving them a greater chance of completing the course successfully.

In those environments that take full advantage of IT’s capabilities, such assessments are incorporated into course software. In its redesign of introductory statistics, Ohio State will integrate, directly into its course software, a learning-style inventory instrument developed by Barbara A. Solomon and Richard M. Felder at North Carolina State University ( This instrument, which helps students develop a self-awareness of their learning style, scores students on their degree of active versus reflective learning, sensing versus intuitive learning, visual versus verbal learning, and sequential versus global learning. The course team will also integrate a study skills assessment instrument, developed by Ohio State’s Academic Learning Lab, to guide students in making appropriate choices from the buffet of learning opportunities.

Riverside Community College’s redesign of its college algebra course is based on using ALEKS (Assessment and LEarning in Knowledge Spaces), a Web-based, artificial-intelligence program that generates individualized student assessments, study plans, and active learning sets. Through sophisticated modeling of each student’s "knowledge state" of elementary algebra, ALEKS focuses clearly and precisely on exactly what the student is most ready to learn at a given moment. Based on this information, ALEKS creates customized active learning sets for each student. Students then work through the customized sets, building momentum, confidence, and ultimately, subject mastery. ALEKS also provides collective reports on the students in all classes, pointing out common problem areas that can be addressed. Because ALEKS is Web-based, it is available to students twenty-four hours a day, seven days a week.

2. An Array of Interactive Materials and Activities
All the new pacesetters offer students a broad array of learning materials and activities. In Virginia Tech’s Math Emporium, for example, numerous types of learning experiences are available. Students gravitate toward the kind of experience they find best for them. Since students have different learning preferences, the availability of recorded lectures and interactive, Web-based materials enables some students to complete the course primarily on their own, interacting with faculty and other students only to the extent required by the course. Others prefer to take advantage of the variety of support activities and facilities as well as opportunities to interact with course faculty, teaching assistants, and peer mentors. By working collaboratively to design the course, faculty members are able to create, change, adapt, and add to an ongoing body of materials.

Effective Web-based materials, often called learningware, go far beyond simply transferring traditional material to the Web, since a simple transfer cannot improve learning. Rather than replacing textbooks, these materials supplement them with activities: interactive simulations that can be actively manipulated, that provide engaging and challenging tasks, and that supply instant feedback on performance. Computer games like "flight simulator" are the ready analogy here; these can be devised in virtually any field. Good learningware engages the full range of the human senses through multimedia technology (e.g., visual examples of concepts, short news clips, or foreign-language conversations that can be reviewed as many times as a student desires) and almost always forces students to make learning decisions. In other words, good learningware encourages active learning.

3. Individualized Study Plans
Unlike traditional course structures that engage students in the same series of activities regardless of students’ disparate abilities and interests, individualized learning environments permit students to move quickly through content they already know and spend more time on areas they find more challenging. Students engage in study at their preferred time rather than at prescheduled times. Students do not all have to do the same thing but rather learn at their own pace.

The new pacesetters’ courses are not completely self-paced, since experience shows that laissez-faire, unstructured, totally self-paced models do not work well and can lead to high attrition rates. Having freedom and responsibility for their own learning may be substantially different from students’ previous educational experiences. The greatest problem is getting students to spend time on task. Some students are extremely slow to log in; if students fall behind, they often lack the support to catch up in time, and many simply won’t make it. Good online programs include a clear structure that paces student learning and builds in mastery assessments to certify progress and achievement of learning goals. Commercial course-management software packages such as WebCT and Blackboard are able to track students’ time on task online. Students need help in adapting to this different style so that they do not mistake freedom of choice for a lack of course requirements.

Drexel uses the term "self-scheduled" rather than "self-paced" in describing its new learning environment. Students can plan their work on a particular module to fit their schedule as long as they complete each module by the end of the week. Thus, at the end of each week, all students working on a particular module will have taken the final assessment for that module and will be at the same point: ready to move on to the next module. The goal is to maximize students’ flexibility in learning the course material as best fits their learning preference and schedule while providing enough structure for them to make the same kind of forward progress as in a traditional course. Linking students to a definite learning plan with specific mastery components and milestones of achievement and creating some form of early-alert intervention system are critical components of course design.

4. Built-in Continuous Assessment
When faculty members shift the traditional periodic assessment model (midterm and final examinations) toward continuous assessment, students view assessment as a learning experience rather than as an all-or-nothing performance measure. Few people would be surprised to learn that students, if allowed to do so, will often put off study until shortly before exams and that such cramming does not lead to long-term retention of information. Spacing quizzes (either graded or non-graded) throughout the semester improves overall understanding and retention of terminology and concepts.

The advantages of continuous assessment include an increase in the time that students spend studying, a higher level of familiarity with tested material and of comfort with the testing process, immediate feedback, and the ability to see the result of effort on achievement. Assessing students’ understanding of concepts is very effective in detecting areas in which students are not grasping the concepts, thereby enabling corrective actions to be taken in a timely manner, and in preparing students for higher-level activities in the computer labs. Periodic mastery testing helps students keep up with the readings and recognize holes in their understanding, and it promotes understanding of the content. Threading assessment continuously throughout a course also obviates the threat of cheating.

Online assessment tools, moreover, have increased in sophistication and now make continuous assessment more feasible and easier to manage. UIUC’s Mallard and Michigan State’s CAPA are two examples of these sophisticated software tools. Computer-adaptive testing and assessment of individual students’ strengths and weaknesses can craft customized paths of learning that present learning materials tailored to meet assessed gaps in abilities and provide tasks that are appropriately challenging. Carnegie Mellon University has developed an "intelligent tutor" that can follow a student’s progress and adapt the learning environment to respond to areas of difficulty a student may have. The ALEKS mathematics software package can quickly display the location of individual learners or groups of learners on a particular vector of development, allowing faculty mentors to plan interventions accordingly.

5. Appropriate, Varied Human Interaction
Helping students feel that they are a part of a learning community is critical to persistence, learning, and satisfaction. In many cases, human contact is necessary for more than just learning content. Encouragement, praise, and assurance that they are on the right learning path are also critical feedback components, helping students get through rough times and keep on working. Knowing that someone is there to help when they get stuck and to get them moving again gives students the confidence that they can succeed.

Such active mentorship can come from a variety of sources, such as traditional instructors (faculty and graduate teaching assistants) and more advanced undergraduate students. Access to a large support system of fellow students and tutors who are available virtually around the clock is a key component to these new designs.

Students also learn from each other. Research has shown that students in distance education take on the role of "teacher" more often than do students in traditional classrooms. This not only has obvious implications for the content and mode of instruction but also sets up a model of learning communities that is invaluable when our students enter the work world. Knowledge-management software can structure a situation in which students can be actively encouraged to get in touch online with others who recently encountered and overcame similar problems.

III. Increasing Access to Higher Education

When asked how online learning can lead to greater access to U.S. higher education, most people think about increasing access to campuses and their current structures and services. Symposium participants were able to come up with far more creative ideas about access—ideas that take advantage of the capabilities of information technology and the Internet. In so doing, they broadened considerably what we mean by access, moving beyond giving students who cannot travel to a classroom the opportunity to participate in higher education. Access means different things to different people; it does not have a one-size-fits-all definition. Information technology enables us to expand our definition of access to meeting the learning needs of diverse students when, where, and how they so desire. Technology can create environments that provide individualized access to learning, access that serves each person in ways that he or she can most benefit.

Symposium participants generally agreed that the key to designing more-accessible learning environments is to eliminate constraints. As one participant put it, the more virtual (anyplace, anytime) the delivery model, the more accessible it is. Too many distributed learning models still burden students with the constraints of time and place (someplace, same-time). In addition to those of time and place, there are academic constraints that contribute equally to limiting access. Just as the standard semester is emblematic of time constraints, so do standard academic structures like the three-credit course or the institutionally based degree program restrict access to higher education.

Asynchronous learning environments have done a lot to eliminate the constraints of time and place, but have they done as much as possible to take advantage of the capabilities of the Internet? The overwhelming majority of online programs, like their on-campus counterparts, follow traditional term (semester or quarter) "class" models, a classic case of applying old solutions to new problems. Why? Most surely the reason is institutional convenience; few would argue that students prefer fixed start times. Clearly, information technology can support new structures that offer greater flexibility for students. Indeed, without the support offered by information technology, individualizing instruction is both expensive and logistically challenging. Once an institution recognizes how information technology can manage a more diverse approach to organizing instruction, there is little reason to retain a lock-step approach.

The Groundbreakers

In contrast to prevailing practice, Rio Salado College, the University of Phoenix, and Cardean University have revolutionized the college calendar. At all three institutions, entering students do not have to wait until the next semester begins in order to enroll. At Rio Salado, students have access to more than two hundred online courses, the majority of which start twenty-six times a year (the remainder usually start six to eight times a year). This means that any student who wants to take a course never has to wait more than two weeks to start. In addition, although each course is advertised as a fourteen-week class, students are allowed to accelerate or decelerate as needed. Rio never cancels a class that is offered online. If only one student enrolls, he or she can be accommodated. Information technology provides the management system that enables faculty members to handle several starts at once, keeping everyone on track.

The University of Phoenix uses a rolling-cohort model in its online programs, enabling a course to begin as soon as eight to thirteen students are ready to start a particular study. Cardean University also allows its MBA students to begin at any time, once a cohort of about twenty-five is established. In both cases, students share a common discussion environment and an instructor, whose role is to build community and facilitate students’ discussion of the application of course concepts to their work environment.

The University of Phoenix, like Rio Salado and Cardean, makes flexible access to its programs and courses one of its highest priorities. Study at each of these institutions is primarily asynchronous. Each has a focus on providing greater access to learning for working adults—the majority of whom are isolated from typical college classrooms by time, geography, or transportation barriers—and each has designed its environment accordingly. When these institutions survey students about what they like most about their programs, the convenience factor of having access to education when students want and need it always ranks high.

The New Pacesetters

Just as information technology enables institutions to create more flexible access to existing courses and programs by eliminating the constraints of time and place, so too does it allow us to expand our definition of access to eliminate academic constraints. Pacesetting institutions are increasing access via information technology in three important academic dimensions: academic resources, degree programs, and learning through modularization.

Increasing Access to Academic Resources
Some institutions are going beyond creating access to tradi-tional faculty and other academic resources. Rio Salado has found a way to eliminate an academic constraint—the need for students to come to the physical campus to take laboratory-based courses. Rio teaches anatomy and physiology courses completely online using virtual techniques, reducing laboratory costs without sacrificing quality (see Rio Salado College, Online Human Anatomy).

Out of several hundred online courses offered by Rio Salado, four science courses rank in the top eleven, an indicator that students want this expanded kind of access. At the British Open University, expensive or dangerous home experiment kits (e.g., chemistry laboratories, telescopes, microscopes) have been replaced by virtual instruments and experiments. Most science courses at the university are CD-ROM-based. In addition, both Rio Salado College and the British Open University emulate field trips online (e.g., field trips by geology students have been replaced by virtual field trips that can draw more easily on high-quality support materials).

Increasing Access to Degree Programs
Whereas most people in higher education think about access within a construct of time and place, symposium participants pointed out that academic policy constraints are frequently more potent. Expanding access to higher education requires overcoming the many academic barriers established by individual institutions. The issues of access to full-degree programs and of credit transfer among multiple institutions have been a challenge to colleges and universities for many years, but the existence of the Internet and the explosive growth of online learning have radically escalated their importance.

Although not all online learners seek degrees, many do. Public-policy goals that drive most virtual university efforts, for example, include increasing the number of degree-program graduates. As one symposium participant commented, it is an accepted truism in higher education that adult learners will not begin a degree program if they cannot see how they will complete it. If online learning is going to expand access significantly in the near future, we will need to increase the number of degree programs that students may complete entirely at a distance.

Institutions that accept transfer credits or work experience relatively freely, while offering virtual degree or certificate programs, are especially effective at increasing access. Many students may prefer to take courses from more than one institution, and this trend is accelerating.

At Excelsior College, students can pick and choose how they will complete their degrees and which learning services they need to advance their educational goals, depending on their particular life circumstances. Some students use all of the Excelsior College learning services, some use none, and some pick and choose. Some students complete their degrees exclusively through credit-by-examination, some take courses from many different institutions, some rely on distance education courses to complete their degrees, and some attend only one institution in their local communities. Some students take a few examinations, a few distance courses, and a few classroom courses taught at one or two local institutions. In addition, since Excelsior places no caps on the kind or amount of transfer credit that it will recognize from regionally accredited colleges, students do not need to repeat learning they have already achieved in other collegiate settings. The college thus reduces the total number and cost of courses that students must take to complete their degrees (see Excelsior College, What You Know Is More Important than Where or How You Learned It).

Students at Excelsior College have neither cohorts nor calendars. They can create programs of study that combine on-campus courses, online courses, test preparation, and independent study to individualize the time and place of study while achieving common learning outcomes as validated by Excelsior’s highly regarded standardized examinations. Trading academic residency for rigorous assessment clearly expands access to higher education.

Increasing Access to Learning through Modularization
To make learning available to the greatest number of people, we can modularize or break down educational content into smaller chunks that can be reorganized or recombined to meet the learning needs of individual students. Modularization characterizes most of the features of the buffet-style courses described in the preceding section on improving quality.

By modularizing course content, pacesetting institutions are able to tailor the study to different types of students with different goals. Like Ohio State, Drexel University is exploring how modularization can benefit both students and institutions (see Drexel University, Modularizing Computer Programming).

Modularization moves us further along the continuum from what has been called just-in-case learning to just-in-time learning. As applications of information technology become more sophisticated, we can identify weaknesses in students’ learning as they progress through a course. Students can then focus on these areas of weakness and spend less time on content areas they already understand. Customized learning materials can be presented to students in order to provide more practice and/or greater variety in the types and levels of difficulty. With such focused study, students can potentially decrease the time they spend on a particular course, increase their success rates, and reduce the number of times they repeat a course, all of which play an important role in increasing access to U.S. higher education.

IV. Reducing the Costs of Teaching and Learning

When the issue of cost is raised in relation to online learning, many people in higher education focus on the question, does online learning cost more or less than traditional instruction? The predominant belief is that it costs more. Temple University’s president, David Adamany, typifies the views of many; he was recently quoted in the Chronicle of Higher Education as saying, "No one has yet found a way for online learning to be economically viable."3

The issue of cost is directly related to that of access. As one symposium participant noted, it is very difficult for most existing institutions to expand access, whether on campus or online, without facing significant budget increases. Without new funding sources, enrollments can only expand on the margin: where courses and programs have insufficient enrollment and new students can fill empty seats. A contributing factor is that productivity in higher education is declining. Between 1977 and 1997, the number of students in higher education has increased by 27 percent while the number of faculty has increased by 56 percent, resulting in a decline in the student/faculty ratio from 16.2:1 to 15:1.4

One symposium participant commented that faculty, via threats of unionization, had forced the participant’s institution to limit the number of students in online courses to twenty, which in turn limits the ability both to scale (i.e., produce more cost-effective courses) and to serve more students (i.e., increase access). Indeed, a new, emerging paradigm for traditional online courses calls for a 20:1 (or less) student/faculty ratio, reflecting the on-campus small seminar. Campus leaders are rightly concerned that such applications of information technology are increasing instructional costs rather than controlling or even reducing them. Online learning offers enormous possibilities for guiding and managing instruction, for communicating with students, and for assessing student performance and knowledge on a much larger scale than is currently the norm if we can change the student/faculty ratio. The issue is, how can we handle large numbers of students cost-effectively?

Rather than simply comparing the costs of one form of instruction with another, symposium participants were asked to consider the following question: What kinds of approaches to online learning do you believe can lead to a reduction in instructional costs? By thinking of ways to take advantage of the capabilities of information technology and the Internet and, in so doing, by reconceptualizing the way that courses are designed, participants were able to come up with many creative ideas about how to make collegiate instruction more cost-effective.

The highest cost component of instruction is faculty personnel. Currently, the job of a faculty member—whether in class or online—is seen as monolithic: a collection of tasks that are, with few exceptions, carried out by one person. Faculty usually believe they must and will play all roles in the course-development and course-delivery process. Traditional online providers suffer from what one symposium participant called a "craft mentality," in which a high-priced faculty member is her or his own developer and technical support person, not to mention learning theorist. Information technology offers the possibility of altering this paradigm. Once the many roles or tasks that a faculty member performs are disaggregated—that is, separated and seen individually—the opportunities for substitution and cost reduction become clearer.

Higher education has known for decades that substituting cheaper labor for more expensive labor reduces instructional costs. The use of graduate teaching assistants, adjunct and part-time faculty, and other instructional personnel has enabled institutions to keep their costs from rising beyond what they are now. The knock has always been that our dependency on part-time faculty reduces the quality of instruction, and anecdotal evidence seems to support that view. The academy, broadly, worries about institutions that rely too heavily on adjunct faculty for two reasons: (1) the academic program may fall into the hands and control of administrators who make decisions based on financial expediency rather than academic quality; and (2) quality assurance may be difficult to maintain, since the academy has neither the infrastructure nor the culture to support a close monitoring of ubiquitous and disenfranchised adjunct faculty.

Both the groundbreakers and the new pacesetters follow a strategy of substituting cheaper labor for more expensive labor and of employing more differentiated kinds of labor in both course development and delivery. What distinguishes their methods from higher education’s historic approaches? First, both types of new providers rely on technology-based, common or centralized development of course structures and course materials, enabling a much tighter level of quality control.

Second, both take advantage of the ability of IT to disaggregate instructional roles to even greater levels of granularity while ensuring overall course coherence. Third, both reduce the duplicative development costs of individual faculty members and enhance the quality of instructional and assessment materials. And fourth, both enable multiple faculty to teach the same material and thus to handle more students.

The Groundbreakers

Originated by the British Open University and replicated with their own twists by the University of Phoenix, the Dallas Community College District, and Cardean University, groundbreaking institutions focus on creating an efficient course-development process and supporting that process with tools that increase efficiency. The model is one in which large, up-front investments are made in single courses, using the best expertise possible in the development team, with the expectation that very large numbers of students will ultimately enroll. In 1999, for example, the British Open University piloted what is now its most successful online course—"You, Your Computer, and the Net"—with 800 students. This year, the course had a total student cohort of some 12,000.

For course delivery, the groundbreaking model employs a relatively small core of full-time faculty to set academic standards, oversee curriculum, establish academic policies including degree requirements, and so on. Part-time, adjunct faculty carry out the bulk of instruction. The University of Phoenix, for example, has 240 full-time faculty and more than 8,000 part-time practitioner faculty members. Rio Salado has 25 permanent faculty and 750 adjunct faculty. Quality control is strong because, in each case, courses are developed and monitored centrally, unlike the adjunct model used by most traditional institutions in which part-timers have relatively free rein to teach as they like.

Despite their gains in cost-effectiveness on many fronts, several of the groundbreaking institutions have created a relatively expensive delivery model by restricting the student/faculty ratio to anywhere from 9:1 at the University of Phoenix to 25:1 at Cardean. To support the smaller ratios, Phoenix charges one-third more tuition for its online courses than for its classroom-based courses. While taking advantage of IT to coordinate course development and to ensure a high level of quality control over course delivery, these institutions have failed to exploit fully IT’s disaggregating capabilities. For examples of how this can be done, we turn next to the new pacesetters.

The New Pacesetters

Encouraged by the Pew Grant Program in Course Redesign, several institutions are pursuing an alternative to large, up-front investments in course development. This model takes advantage of existing materials that have been developed commercially or by other universities. In its online college algebra courses, Rio Salado College, for example, requires students to purchase Academic Systems mathematics software just as they would purchase textbooks. Rio then uses this commercially produced software as the foundation for its online mathematics courses. In addition to defraying the cost of materials development, basing the course design on sophisticated software enables instructors to handle a higher number of students (from 25–30 to 125) in their courses, thus further reducing the overall cost per student.

Like the groundbreakers, the new pacesetters reduce course-delivery costs by using technology to serve large numbers of students. Their efforts are differentiated by the further disaggregation of the faculty role and the substitution of technology-based interactions for human labor. Though appearing more traditional than the groundbreaker model in many ways—especially since full-time, tenured faculty frequently serve as lead faculty in course delivery—the new pacesetter model is, in fact, more radical and thus offers greater possibilities for both cost savings and quality improvements.
A straightforward example of this approach is how the University of Illinois at Urbana-Champaign (UIUC) has doubled enrollment in foreign language courses by relying heavily on Mallard, a UIUC-developed intelligent assessment software program that automates the grading of homework exercises and quizzes (see University of Illinois at Urbana-Champaign, The Spanish Project).

Pacesetting institutions are breaking through the small-seminar model for online instruction and are creating new paradigms that are both high-quality and cost-effective. Once again, individualization is the key idea. Our buffet metaphor is appropriate here. Rather than serving a "fixed meal" of instructional resources, these new designs allow students to take advantage of resources according to their own needs. Redesign involves moving from an expensive and inefficient push strategy, which presents all material to all students in the same way and at the same time regardless of their particular needs, to a pull strategy. Students access the material they need when they need it, an approach that takes into account differences in learning preferences and abilities. The latter strategy is not only more effective in dealing with learning issues but also more economical in dealing with resource issues because students use only as much resource as they need. Organized around computer-based assignments, with on-demand tutorial assistance provided as required, these new designs are dramatically reducing both student failure rates and instructional costs.

High-cost, full-time faculty members are no longer the only resource. Instead, resources are matched to the level of difficulty and type of instructional task. Different types of personnel are employed to do different kinds of tasks. In its redesign of its college algebra course, Rio Salado, for example, has found that 90 percent of students’ questions were not math-related and did not require a faculty member to respond. Rio hired an aide to answer these questions, leaving the faculty member free to respond to content-related questions and consequently to handle more students. Possible substitutions used in pacesetting courses include nontenured for tenured faculty, adjuncts for full-time faculty, graduate teaching assistants for various kinds of faculty, undergraduate teaching assistants for faculty or for graduate teaching assistants, and professional staff for traditional faculty.

As an example, Virginia Tech has redesigned its linear algebra course, taken each year by 2,000 first-year students majoring in engineering, physical science, and mathematics. Virginia Tech, like most other higher education institutions, tried to control costs in the traditional mode by employing a mix of tenure-track faculty (ten), instructors (thirteen), and graduate teaching assistants (fifteen) to teach thirty-eight sections of the course. The redesign radically changed the mix of human and technological resources, resulting in a two-thirds reduction in the cost per student (see Virginia Polytechnic Institute and State University, The Math Emporium: Student-paced Mathematics 24x7).

Although many believe that learning environments targeted to particular learning styles and individual learning needs are more expensive than traditional one-size-fits-all methodologies, the introduction of new designs based on information technology can allow for more cost-effective ways of learning—cost-effective for both the institution and the student. As noted above, the new pacesetters’ buffet-style courses have five key features that can improve the quality of student learning. These five features are also major contributors to cost reduction.

1. Assessment of Knowledge/Skill Level and Learning Style
A first step in implementing a pull strategy in which students use as much instructional resource as they need is to assess their knowledge and skill level as they enter the course or program and determine their preferred learning style. Based on those assessments, students can then elect the most efficient path through the required course materials. Drexel’s modular approach to its introductory computer programming course, for example, allows students to earn from one to three credits based on their performance on a knowledge and skills placement test. Students do not need to spend time covering material they already know and can move on to other studies. Drexel can reduce the amount of instructional resources to correspond more accurately to students’ needs. Similarly, Ohio State’s modular format will enable it to eliminate one-fourth of the course repetitions, thereby opening slots for an additional 150 students per year.

2. An Array of Interactive Materials and Activities
Each of these new learning environments reduces the number of lectures and/or class meetings, replacing presentations of content with a variety of activities supported by interactive software. Some eliminate several lectures; others eliminate all lectures. The premise is that faculty do not need to spend as much time (or any time) presenting information. Lectures are replaced with a variety of learning resources, all of which involve more-active forms of student learning or more-individualized assistance. In many instances, computer-based tutorials and feedback substitute for instructor-based tutorials and feedback. Such a strategy is not only more effective in dealing with learning issues but also more economical in dealing with resource issues because students use only as much resource as they need. Savings occur from reducing the number of instructors required and also from freeing up classroom space. Reducing classroom contact hours, for example, from three to one or two through the use of virtual instruction makes it possible for up to three courses to use the classroom hours previously reserved for one class.

3. Individualized Study Plans
Without the availability of information technology tools, creating and managing individualized study plans for students would be highly labor-intensive and hence costly. Sophisticated course-management software, however, enables faculty to monitor students’ performance, track students’ time on task and overall progress, and intervene when necessary to correct a student’s deviation from planned study on an individualized basis. Students can create a definite learning plan requiring periodic log-ins (e.g., students have to take a quiz by—not at!—a fixed time every week and an exam by a scheduled date at the end of each module). Many types of communication can be automatically generated to provide needed information to students. Instructors can use e-mail to communicate with students as a way to encourage students to "come to class" with online materials. Regular weekly, computer-generated e-mails can inform students about their progress and, if necessary, suggest additional activities based on homework
and quiz performance.

4. Built-in Continuous Assessment
The automated grading of homework (exercises, problems), low-stakes quizzes, and tests and exams for those subjects that have correct or easily assessed outcomes not only increases the level of student feedback but also offloads these rote activities from faculty and other instructional personnel. The result is either a reduction in the number of required instructors or the ability to increase the number of students in any given course. Michigan State has shown that the application of technology can reduce the instructional costs of large traditional lecture courses from 10 percent to 30 percent. The largest cost savings was due to the reduced need for teaching assistants for grading and recitation sections.

5. Appropriate, Varied Human Interaction
Faculty who teach traditional online courses frequently complain about overload due to the difficulty of responding to numerous e-mails or managing complicated listservs. The best of today’s threaded discussion technologies enable easy-to-access and easy-to-manage communication among students and between students and their instructors. Wise instructors may seed class-wide discussions and monitor these discussions, but they seldom take responsibility for responding to every posting by a student. They emphasize student-to-student interaction and interaction with the material in ways that force students to formulate most of their postings for peer review and response by their fellow students. Instructors who use these technologies and pedagogies ask students to take more responsibility for their own learning. By emphasizing student-to-student mentorship and interaction as much as possible, we can increase student involvement and improve learning outcomes. This not only is effective but also saves expensive faculty time.

V. Sustaining Innovation

Throughout this paper, we have reiterated the view that individualization is the key to moving beyond the "no significant difference" phenomenon. Currently in higher education, both on campus and online, we individualize faculty practice (that is, we allow individual faculty members great latitude in course development and delivery) and standardize the student learning experience (that is, we treat all students in a course as if their learning needs, interests, and abilities are the same). The conclusion reached by symposium participants is that we need to do just the opposite: individualize student learning and standardize faculty practice.

It is curious that most academics react with horror at the thought of standardizing faculty practice but do not think twice about standardizing the student learning experience. With its connotations of words like regulate, regiment, and homogenize, the word standardize does not precisely capture what we mean. What we need is greater consistency in academic practice that builds on our accumulated knowledge about improving quality, increasing access, and reducing costs. Sustaining innovation depends on a commitment to collaborative development and continuous quality improvement that systematically incorporates feedback from all involved in the teaching and learning process.

The Internet offers unprecedented opportunities to collect, organize, and analyze large, real-time research. Online environments provide enormous information-capturing potential because every move that every student and every faculty member makes is potentially recoverable and able to be analyzed. Sources include responses to online surveys regarding student satisfaction and perceptions; tracking of learner behavior on site (On what learning points do students spend the most time? What is the sequence and pattern of interest? What questions do students ask?); transactional data on student registrations, dropouts, and completions; and interaction and outcome data generated from baseline assessments, exercises, and exams.

To take advantage of these capabilities, we need a new kind of "institutional research" designed to determine which are the most efficient and effective paths for different kinds of learners in particular curricula or courses, so that we can make active adjustments in learning designs. We also need to be much more sophisticated about monitoring and measuring costs. Students, instructors, institutions, accreditors, and consumer agencies all have access to this data, enabling benchmarking and competency assessment. Because of the feedback available, digital products and services can be fine-tuned, and product development can be accelerated. The ultimate vision here is the kind of continuous quality improvement systems used by automated industrial production systems that are for the most part self-monitoring.

It is not coincidental that the new providers discussed above have taken the first steps toward implementing this vision. At the institutional level, Excelsior College, Rio Salado College, the University of Phoenix, and the British Open University are known for building a continuous assessment loop through the collection, analysis, and dissemination of data. In monitoring the quality and effectiveness of its academic program—the strengths and weaknesses of the materials and services provided—each keeps an eye primarily on two things: student learning outcomes and customer and student satisfaction with all experiences at the institution. Excelsior College, for example, does a major student-satisfaction survey every three years. For each graduate, the college does a six-month follow-up survey and a three-year follow-up, as well as an additional three-year follow-up for students who complete graduate school. The University of Phoenix conducts end-of-course surveys among both students and faculty in order to gauge the success of both the individual class and the individual instructor. The British Open University tests and edits its courses based on assessment data that is collected throughout the course-development process.

At the course level, Virginia Tech, Michigan State, the University of Illinois at Urbana-Champaign, and all of the projects involved in the Pew Grant Program in Course Redesign treat the course not as a "one-off" but as a set of products and services that can be continuously worked on and improved. Two factors in their design strategies are key: the collective commitment of all faculty teaching the course and the capabilities provided by information technology. Would it be possible for a single professor conducting an online class to develop such creative, comprehensive, learner-centered designs as exemplified by the new pacesetters? Perhaps, if the individual spent most of his or her career working on the class. Would it be possible for institutions to offer a buffet of learning opportunities to thousands of students annually without the aid of information technology? Most certainly not. IT enables best practices to be captured in the form of interactive Web-based materials and sophisticated course-management software. Rather than reinventing the wheel at the start of each term, the new pacesetters can add to, replace, correct, and improve an ever-growing, ever-improving body of learning materials. This, in turn, leads to greater possibilities for individualization.

Earlier in this paper, we commented that the leading institutions described in the cases do not offer full-blown solutions to the question of how to move beyond the "no significant difference" phenomenon but instead illustrate pieces of the puzzle. Because they share a commitment to continuous quality improvement, all are in an excellent position to incorporate ideas from others. Already committed to a rolling-cohort strategy, the University of Phoenix, for example, could enrich its approach by assessing students’ learning styles, creating cohorts based on those assessments (either homogeneous or heterogeneous), and designing course variations to correspond accordingly. Virginia Tech’s math courses and UIUC’s foreign language courses could incorporate the credit and content modularization ideas pioneered by Ohio State and Drexel. Groundbreakers in distance learning, Rio Salado and the British Open University could learn from the on-campus buffet providers and set new standards of excellence for off-campus learners. In each case, the systemic approach of the new providers enables them to incorporate the best of online academic practice.

This symposium was the fourth of the Pew Symposia in Learning and Technology. The purpose of this symposia series is to conduct an ongoing national conversation about issues related to the intersection of technology and student learning and ways to achieve this learning cost-effectively. The new providers who participated and others cited in the paper are creating a new higher education paradigm, which includes new boundaries for behavior, new guides to action, and new rules for success. As we continue to develop online courses and programs, let’s follow their lead, building on the strengths of the Internet to create new learning environments that surpass traditional modes of instruction.


  1. Charles D. Dziuban, Patsy D. Moskal, and Emily K. Dziuban, "Reactive Behavior Patterns Go Online," Journal of Staff, Program, and Organizational Development 17, no. 3 (fall 2000): 171–82.
  2. The Pew Grant Program in Course Redesign is a three-year, $6 million program conducted by the Center for Academic Transformation at Rensselaer Polytechnic Institute with support from the Pew Charitable Trusts. The purpose of this institutional grant program is to encourage colleges and universities to redesign their instructional approaches using technology to achieve cost savings as well as quality enhancements. The program is supporting 30 large-scale redesigns that focus on large-enrollment, introductory courses and that have the potential to influence significant numbers of students and to generate substantial cost savings. For complete information about the program, including individual project descriptions and cost savings data, please see
  3. Goldie Blumenstyk, "Temple U. Shuts Down For-Profit Distance-Education Company," Chronicle of Higher Education, July 20, 2001.
  4. Thomas D. Snyder and Charlene M. Hoffman, Digest of Education Statistics, 2000 (Washington, D.C.: U.S. Department of Education, National Center for Education Statistics, 2001), 193–94, (accessed October 4, 2001).

Symposium Participants

John Arle
Faculty Chair, Science Department
Rio Salado College

George Connick
Distance Education Publications

Charles D. Dziuban
Director, Research Initiative for Teaching Effectiveness
University of Central Florida

Leigh S. Estabrook
Dean, Library and Information Science
University of Illinois at Urbana-Champaign

Peter Ewell
Senior Associate

William H. Graves
Chairman and Founder

Joel M. Greenberg
Director, Interactive Multimedia
British Open University

Carolyn G. Jarmon
Associate Director
Center for Academic Transformation

Jorge Klor de Alva
President and CEO
Apollo International

Robert F. Olin
Dean, College of Arts and Sciences
University of Alabama

Paula E. Peinovich
Vice President for Academic Affairs
Excelsior College

Pamela K. Quinn
Assistant Chancellor, Educational Telecommunications
Dallas County Community College District

Carol Scarafiotti
Dean of Instruction
Rio Salado College

Kurt A. Slobodzian
Vice President, Instructional Technology
University of Phoenix

Michael R. Thoennessen
Professor, Physics and Astronomy
Michigan State University

Carol A. Twigg
Executive Director
Center for for Academic Transformation

Virtual Participants

Thomas M. Duffy
Cardean University

Nira Herrmann
Head, Department of Mathematics and Computer Science
Drexel University

Diane Musumeci
Associate Professor of Italian, Spanish, and Portugese
University of Illinois at Urbana-Champaign

Dennis K. Pearl
Professor of Statistics
The Ohio State University


Patricia Bartscherer
Program Manager
Center for Academic Transformation

Innovations in Online Learning: Moving Beyond No Significant Difference, by Carol A. Twigg

© The Pew Learning and Technology Program 2001
Sponsored by a grant from the Pew Charitable Trusts.

Center for Academic Transformation, Rensselaer Polytechnic Institute
Dean’s Suite, Pittsburgh Building
110 8th Street, Troy, NY 12180
518-276-6519 (voice)
518-695-5633 (fax)

Return to Top of Monograph