"Does Using Technology in Instruction Enhance Learning? 
or, The Artless State of Comparative Research" 
 
The Technology Source, "Commentary,"  June 1998
http://horizon.unc.edu/ts/commentary/1998-06.asp

Jerald Schutte, a Sociology Professor at CSU-Northridge, has attracted a lot of attention for his 1996 study on the use of technology in instruction. Schutte reported that students in his "virtual" class performed 20% better than students in his "traditional" class, and the results of the study were widely disseminated on listservs and the popular press (including piece in the Chronicle of Higher Education, Feb. 21, 1997, A23). His article, Virtual Teaching in Higher Education: The New Intellectual Superhighway or Just Another Traffic Jam? is available online at: http://www.csun.edu/sociology/virexp.htm 
 
Unfortunately, Schutte's research design and methodology are so flawed that the results of the study are uninterpretable. His fundamental error arises from a confusion between teaching methods and delivery systems, an error which is common to many studies in this field. 
 
The Study 
 
Students enrolling in Schutte's course in Social Statistics (n = 40) were randomly assigned to a "traditional" class or a "virtual" class in roughly equal numbers. Students in the traditional class (where he presumably lectured) met every Saturday during the 14 week semester and submitted weekly problem assignments. By contrast, here is Schutte's description of the virtual class: 
 
The virtual class had four assignments each week: 1) e-mail collaboration among randomly assigned groups of three students in which they generated weekly statistical reports and sent them to the instructor using e-mail; 2) hypernews discussion in which a weekly discussion topic was responded to twice a week by each student; 3) forms input via the WWW which allowed for student submission of the same homework problems being solved by the traditional class; and 4) a weekly moderated Internet relay chat (mIRC) in which student discussion and dialogue were carried out in real time in the virtual presence of the professor. 
 
At the end of the semester, the virtual students scored an average of 20 points higher on the 100 point midterm and final exams, which Schutte attributes to "virtual" interaction: 
 
This experiment was intended to assess the merits of a traditional, versus virtual, classroom environment on student test performance and student affect toward the experience. It was hypothesized that face-to-face professor-student interaction is crucial to test performance. However, the data indicate the reverse, that virtual interaction produces better results. (Emphasis added.) 
 
Schutte goes on to suggest that perhaps the reason the students in the virtual class performed better is because they compensated for the lack of face-to-face contact with the professor by forming study groups to "pick up the slack of not having a real classroom." (He offers no documentation about these study groups, so his evidence is apparently anecdotal.) What he apparently fails to understand is that his study compares students working in groups with students who are not working in groups, regardless of whether they were in self-created study groups or in virtual discussion groups. 
 
Weaknesses in the Study 
 
Actually, students in the virtual class experienced a completely different method of teaching from those in the traditional class. Not only did they have more opportunities to be involved with each other and with the teacher, but (very significantly) they were intensively engaged with the course material over the entire week. It seems clear that Schutte would have to provide similar small-group activities, discussion opportunities, and other assignments for the traditional class if he wanted to test the comparative effectiveness of virtual instruction with traditional instruction. Schutte's study simply demonstrates that particular teaching methods (e.g., cooperative learning and exercises that insure more time on task with the material) yield improved performance, results that educational researchers established years ago. (See, for example, research summaries in Johnson & Johnson, 1989, Nilson, L. B. 1998, and Menges, Weimer, & Associates, 1996.) 
 
There are also a number of methodological weaknesses in this study. Although Schutte did administer a pre-test to measure the demographic similarity of the two groups, he didn't test them for statistical knowledge or aptitude (only "statistical feelings" and "math feelings" on a scale from one to ten). Pre-existing differences in statistical expertise could explain the superior performance of the virtual class, especially given the small number of students in each group. (Indeed, if the virtual group did have more expertise or aptitude, one might expect them to out-perform the traditional class by more than 20%.) Furthermore, Schutte provides only scant information about the course tests and doesn't describe the grading procedure at all. He does say that both mid-term and final exams had four parts 1) matching, 2) objective, 3) definitions; and 4) problems. It would be important to know if the tests were "blind graded," across students as well as across the two groups, especially since portions of the tests require subjective judgment on the part of the grader (definitions and problems). An effective research design would include pre-test and post-test measures and selection of significance levels prior to collecting the data. 
 
In short, the Schutte study violates so many of the rules of empirical research that we cannot use it to draw any conclusions about the comparative effectivess of "real" versus "virtual" teaching. Can we learn anything useful about teaching with technology from his work? One could argue that, without technology, Schutte would not have experimented with other teaching methods. We have all heard anecdotes about professors who, in trying to incorporate technology into their courses, adopt better instructional approaches as a consequence. However, we don't know how many teachers benefit in this way nor whether students in their classes actually learn more material or master more skills. This argument also fails to address the question of whether we need the technology at all if we can be more effective by simply adopting these teaching methods in the traditional classroom--and it would certainly be cheaper. 
 
Alternative Outcomes 
 
Does Shutte's study demonstrate that it is possible to conduct cooperative learning in a virtual environment? The answer to this question depends on one's definition of cooperative learning. At its simplest level, cooperative learning is "a structured, systematic instructional strategy in which small groups work together toward a common goal" (Cooper and Mueck, 1990, p. 68), but authorities on the subject insist that effective cooperative learning requires more than simply putting students into small groups and giving them a task to perform. For example, there must be "positive interdependence" among group members, which means that each member's success must depend to some degree on the group's success. "Individual accountability" is also necessary, which means that all members of a group are accountable for their own learning as well as for the other students in the group, and group members should share the work load equally. The group must also be aware of (and consciously work on) collaborative social skills such as conflict management and giving constructive feedback. The social aspects of cooperative learning groups, which are crucial to their effectiveness, are difficult (if not impossible) to replicate in a virtual environment. Schutte obviously did not set out to create cooperative learning groups by this definition, but his study shows that some aspects of student collaboration can be accomplished in "virtual space." However, even this conclusion is provisional, and it begs the question of how much more effective the groups could have been if they had been "real" instead of "virtual." 
 
Finally, one might look at Schutte's project in terms of the effectiveness of "total instructional redesign," since his methods in the virtual class were so different from his methods in the traditional class. If we accept the results of his study as valid (which, for the reasons given above, we cannot do), his "redesign" could be called effective. However, we cannot ignore the enormous costs of the technology in this equation. If he had used these methods in his traditional class, costs would not increase, but because he and his students needed the networked technology of a major educational institution, the costs of this change are extremely high. One could argue that traditional instruction requires buildings, desks, heat, light, and other resources that cost enormous amounts. However, buildings don't need to be rebuilt every five years and desks don't become obsolete because someone has upgraded the design, nor do traditional classrooms require a platoon of highly-paid experts to stay "on-line." 
 
Other Comparative Research 
 
Schutte's study is flawed, but many of the research studies in this area are also weak. Thomas Russellís survey of research on what he characterizes as "The ëNo Significant Differenceí Phenomenon" (http://tenb.nbcc.nb.ca/phenom/) is often used to support the argument that it doesnít matter what delivery system is used, since there is no difference in how students perform. But if one actually reads the studies on Russell's list, it is obvious that they suffer from many shortcomings: the research designs are poorly conceived, the statistical analysis is weak or absent, and/or the sample size is too small. Many studies donít try to measure learning outcomes at all, but focus instead on attitudinal outcomes--how the students felt about the experience rather than what they learned. Attitudinal outcomes are important, but certainly only in conjunction with cognitive outcomes. The studies that do try to assess student learning as an outcome variable often use tests that measure simple recall of information rather than mastery of higher-order learning. In these cases, it is not surprising that there is "no significant difference" in performance because it really doesnít matter to students how they acquire factual information. 
 
Conclusion 
 
Educational researchers face many difficulties in trying to conduct controlled studies in university settings, since threats to validity and reliability are often beyond the influence of the investigator. As a result, a number of people who are interested in the application of technology to teaching have abandoned traditional research models, opting instead for "proof in practice." The Flashlight Project of the American Association for Higher Education is an example of this approach (a description of the project is available on the AAHE web site: http://www.aahe.org/). This strategy can work, and might ultimately identify successful applications of technology in teaching, as long as evaluations of the results are based on learning outcomes. Unfortunately, most of these projects repeat the same error that Schutte committed, confusing teaching processes with instructional outcomes. 
 
For example, the Flashlight Project seeks to discover "whether faculty and students find the available technology useful (or a hindrance) when they try to implement each of 'seven principles of good practice in undergraduate education.'" The seven principles (from Chickering and Gamson, 1987) are: 
 
Interaction between the student and teacher (or tutor, or other expert); 
Student-student interaction; 
Active learning; 
Time on task; 
Rich, rapid feedback; 
High expectations of the student's ability to learn; and 
Respect for different talents, ways of learning. 
 
Their justification for using this approach reveals an interesting, if flawed, line of reasoning: "Because so much research indicates that these practices support better learning, it would be significant to discover that they were being implemented and that technology was playing an important role." In short, if we find that these secondary indicators are present in technology-based instruction, we should assume that learning occurred. Most teachers know that any teaching method can be performed well or poorly, so the presence of (for example) active learning techniques will tell us nothing about the quality of the learning outcomes. Nor will this project tell us what the students learned (facts, ability to apply knowledge, critical thinking skills) or whether they could have learned better or faster without the technology. The evaluation does not address cost-benefit issues with respect to learning outcomes, so the question of comparative costs of "delivering" instruction with and without technology will go unanswered. These are the kinds of questions that are essential for evaluating any educational program, but especially programs that place increased demands on the human and financial resources of higher education. 
 
In the 1960ís, instructional television promised to change teaching and learning dramatically. College administrators and state legislators, hoping to "expand educational opportunity" (and ultimately save money) by using this electronic delivery system, invested millions in closed-circuit systems, TV production facilities, educational television stations, and even airborne broadcasting systems. Proponents of ITV in the '60s predicted that "as much as 50 percent of the college degree program will be available for credit via television" in the future (Murphey & Gross, 1966, pp. 83-95). Instructional television failed to achieve the transformation of higher education, but the 1960's were boom years for education and the grand experiment with ITV (wasteful as it was) occurred in a time of flush resources. Today, higher education operates in a much tighter fiscal environment and we cannot afford to make many mistakes. We therefore need better studies, focused on significant questions, to guide us in developing appropriate and cost-effective applications of the new technology. To make educational policy decisions or base large investments on anything less is foolhardy. 
 
____________________ 
 
Chickering, A. and Gamson, Z. (1987) Seven principles of good practice in undergraduate education, AAHE Bulletin (March). 
 
Cooper, J. & Mueck, R. (1990). Student involvement in learning: Cooperative learning and college instruction. Journal on Excellence in College Teaching, 1, 68-76. 
 
Johnson, D. W. & Johnson, R. T. (1989). Cooperation and competition: Theory and research. Edina, MN: Interaction. 
 
Nilson, L. B. (1998). Teaching at its best: A research-based resource for college instructors. Bolton, MA: Anker. 
 
Menges, R. J., M. Weimer, & Associates. (1996). Teaching on solid ground: Using scholarship to improve practice. San Francisco: Jossey-Bass. 
 
Murphey, Judith, & Gross, Ronald. (1966). Learning by Television. New York: The Fund for the Advancement of Education.