distance learning courses which use two-way compressed video. With these ... all information comes through the television system. Students at both sites also ...
Using Computer Assisted Collaborative Learning To Improve Distance Education N. A. Pendergrass Department of Electrical and Computer Engineering University of Massachusetts Dartmouth Gregory Sun Engineering Program University of Massachusetts Boston
Abstract - This paper describes how computer assisted, collaborative learning techniques can be used to improve distance learning courses which use two-way compressed video. With these methods, students spend most of their class time working in groups on problems or interactive demonstrations. The approach provides an active learning experience and sharply reduces the amount of information which must be sent through the video system. The methods also increase interaction between students and the instructor and raise student motivation to overcome obstacles in the course. This paper also discusses many of the practical issues that are involved in distance education for classes using either collaborative learning or lecture based methods.
Introduction The two introductory linear systems courses described in this paper are being provided over two-way compressed video from UMass Dartmouth to UMass Boston. The first course is required for majors in electrical as well as computer engineering and it introduces students to important systems concepts and methods of analysis including Fourier, Laplace and Z-transforms. The sequel course is required only for electrical engineering students and introduces digital filtering, feedback control and state variables. Both courses traditionally have been taught as lecture courses without any laboratory or hands-on activity. We were concerned that the addition of a distance education component to these courses could reduce the likelihood of success in the course for some students. For example, after the novelty of the two-way video system wears off, undergraduates at the remote site have to make extra effort to pay attention during a lecture because nearly all information comes through the television system. Students at both sites also have to contend with the technical constraints that the video system imposes on the instructor and them. For example, the instructor has to be sure to stay on camera and cannot move about as freely during a lecture. The video system has to be managed well while also monitoring if students are engaged and following at both sites. In this way the video system weakens the feedback loop between students and the instructor at both sites. During a lecture for a distance learning class, even with two-way video
and audio, it is more difficult to motivate students and to catch, and quickly correct, misunderstandings or faulty thinking. Collaborative learning techniques were chosen to be used in the linear systems courses because these methods have been shown to improve student motivation and performance over traditional lecture approaches [1, 2]. In addition, these methods reduce the amount of information that must be obtained directly from the instructor. Students learn much of the material from each other while working on problems in class. Since most class time is spent on group problem solving, the video system has less impact at both ends and the dependence of remote students on the video system is greatly reduced. Computer assisted interactive demonstrations and activities were integrated into both courses. These are described in detail in [3]. The combination of methods used, as well as the basic design of the classroom, were modeled after those used in RPI’s studio physics classes [4]. They were also heavily influenced by the large amount of literature on cooperative learning and active learning such as found in [1, 2]. The instructor at UMass Dartmouth had not previously used this combination of methods so the linear systems courses were heavily revised and then piloted in the fall of 1995. The pilot versions of both courses were done for a year without a distance component as described in [5]. This provided an opportunity to develop in-class activities and to experiment with the teaching methods before involving remote students.
The Courses The first of the linear systems courses included remote students in September, 1996. There were four students at UMass Boston and 41 students at UMass Dartmouth. The sequel started in January, 1997, with 25 students in Dartmouth and 4 in Boston. TA’s were assigned to the classes at both ends. Picturetel systems were used for two-way compressed video over three ISDN telephone lines. The system at the live end had two large video monitors and a large screen projected display. Video inputs to the system included two remote controlled cameras (one on the instructor, one on the
class), a document camera and a converter for display of the instructor’s computer screen. The system was controlled by a Picturetel podium using a touch screen monitor which also showed the image being transmitted by both sites. Further details of the classroom design, complete with pictures, can be found in [5]. The remote classroom was set up for video conferencing. It had a single remote controlled camera mounted on top of a large monitor. Control of the system could be done through the instructor’s podium or by students at the remote site with a small keypad. Computers were located near the conference table. The courses met twice a week for an hour and fifteen minutes. Usually the classes began with a short reading quiz over the assigned material. This was done to encourage preparation. After collecting the quiz, students were polled for their answers. Areas where students had poor understanding were subjects for discussion. Then a short presentation was done by the instructor to provide a context for the problems and activities that followed. Most class time was spent on problems. On shorter ones, students were told to try to come up with a solution by themselves in some small period of time. At the end of that time they were told to take a similar amount of time to compare results with students nearby and to come to a consensus on a final solution. Longer in-class problems were assigned to informal teams made up of groups of students sitting near each other. In either case, a student would be randomly chosen when the problem was completed to give an answer and explain it. If the result was incorrect or if there was another way to solve the problem, another student was recruited to supply insight. Approximately a quarter of the classes involved some kind of computer problem or activity. The computer was used to accelerate the solution process, to improve visualization or to improve the connection between reality and the concepts of the course. Audio input and output also were often used in computer activities to further help develop student’s intuition about signal processing applications as described in [3].
Important Practical Issues The Document Camera: A document camera was part of the two-way video system at both ends. It was of enormous value in a collaborative learning classroom. It made student presentations of handwritten results very quick and convenient. With the zoom lens, almost any hand drawing or equation was presentable so there was no need to rewrite the solution onto a blackboard or plastic sheet. At either end of the video link, students simply put their solutions under the camera, put on a microphone, and described their path to a solution. Display Resolution: Two-way video systems are usually limited to ordinary TV resolution. Computer monitors routinely have much better resolution so that graphs, line drawings, block diagrams and even presentation material from a computer can be unreadable when crossing
any interface to the video system unless line widths and font sizes are chosen to be extra large. This was a serious limitation for computer demonstrations by video link to the remote site. Fortunately, most computer activities in the collaborative learning classes were conducted by students on their local workstations with high resolution displays. However, there were occasions when we wanted to share a particularly creative student result with all students but could not. To avoid having to send a computer screen over the video system, we are exploring the use of Internet application sharing software such as Microsoft NetMeeting. This would allow one screen to be displayed on all other screens at both sites. This has been successfully done between computers at UMass Dartmouth and will be tried with the remote classroom computers as soon as they are connected to the network. Monitors, Cameras and Eye Contact: Placing a camera directly above the monitor develops the illusion of eye contact in the same way a TelePrompTer works for the evening news on TV. By placing the camera and monitor part way back in the room and at the side, the instructor was able to keep track of student activity at the remote site and keep eye contact with them. Preparing For Loss Of Connection: There is rarely any warning when it occurs. The video system simply disconnects and the screens go black. About one in eight classes of an hour and a half long have had some such problem between our sites. The most frequent cause seemed to be sensitivity of the video compression system to anomalies in the ISDN data lines. However, several failures were caused by faculty who taught other classes and who did not understand how to use the system and left it in a strange situation. Fortunately, problems longer than ten minutes with video connections during class have been rare. In addition, during collaborative learning exercises the video system was not very important and nobody noticed immediately when the system went down. If there was a difficulty in getting the video connection reestablished, a telephone call to a speaker phone at the remote class was used to answer questions and launch students into the next activity while attempts were made to reconnect. This was easier because handouts for the class were routinely faxed and copied for students the day before. In addition, there were fax machines in both classrooms. Office Hours: Collaborative learning methods dramatically reduced the number of student questions for the instructor and the TA’s outside of class at both sites. The methods appear to better prepare students during class to be successful on homework problems. In addition, cultivation of a cooperative atmosphere has resulted in much more use of peers for help. For these reasons, remote video or telephone office hours proved unnecessary. When we had them they were not used. As an alternative, the video system was left on after class to allow casual interaction with the instructor or other students.
Moving Paper Back And Forth: The only real aggravation in teaching the course has come from trying to get student work reliably delivered between sites in a timely way. We would like to be able to get exams graded and returned in one week or less and homework in ten days. We have tried faxing homework and exams for grading but pencil marks do not reliably copy or fax well. There is a daily courier between campuses but the normal process still involves departmental offices and campus mail at both sites. While next day delivery is theoretically possible, in practice it has been more like two to four days one way. The Personal Touch: We have found that a two-way video system works well with collaborative learning methods to maintain good personal relationships with students; however, it is hard to start them over a video link. For this reason the instructor made a trip to meet remote site students within the first two weeks of class and again shortly after mid semester in each course. Controlling Computers During Class: The computer is an irresistible distraction. Games and the Internet are just a few examples of marvels vying for student attention during class. Classroom design was used to help students resist temptation. For example, we arranged the computers so that students had to turn away from the keyboard and display during mini-lectures. We also placed the monitors at both sites so the screens were easily visible to the instructor. This also made it easy for the instructor to determine how each group was progressing during a computer activity. We learned quickly that computer problems in class could take unacceptable amounts of time if they were not carefully set up. For this reason we chose in-class activities which avoided any programming. Students typically were asked to change parameters in MATLAB programs that we provided and to explain and understand the resulting outcome. They also manipulated SIMULINK block diagrams using a limited set of well chosen blocks to achieve a desired system result. Even moderate programming activities in MATLAB have been confined to team project assignments outside of class. Reference [3] has descriptions of some example interactive computer activities using MATLAB and SIMULINK for the linear systems courses. The M-files and other descriptive material are available on the Internet at: http://www.ece.umassd.edu/ece/tour/lin_web.htm.
Results The combination of collaborative learning methods and computer assisted, hands-on activities have proven to be very effective in the sequence of linear systems courses at both the live and remote sites. Student motivation in the classes appeared to be high. They obviously liked the class format and commented favorably about it. In addition, students often stayed after class to experiment with a computer based activity. Students at the remote site appeared to like the course format over the video system. The sequel course was not
planned to be offered in Boston but the students there recruited an additional student and campaigned successfully to have the second course offered. A very low drop rate was experienced with the distance courses as well as with the pilots that did not include a distance component. This appeared to be a result of using the computer assisted collaborative learning methods. Students were apparently better motivated and could find ways to learn the material and succeed in these courses. Only one student dropped the class for a non-medical reason at UMass Dartmouth in each semester. Four students started the first course in Boston and one of those disappeared for unknown reasons long before the first exam. In the second course, none dropped in Boston. In an exit survey of students at UMass Dartmouth after the first distance course, 18% of the students said that they would have preferred more lecturing and 34% said they learned more from lectures than from working problems during class. After the sequel course where most students had seen the new methods for two semesters, only 5% would have preferred more lecturing and just 6% reported that they learned more from lectures than from working problems in class. 95% of students in both semesters said they preferred working in groups of two or more and 96% indicated that computer projects improved their learning. 85% of those in the first course and 76% of those in the second course indicated that they spent more time on signals and systems than they did on any other course that semester. These results are similar to those from the pilot courses without remote students [5]. Most students at the live end of the video link seemed to enjoy the addition of remote students to the class. The majority of students at the remote site in Boston were somewhat older than the students at UMass Dartmouth and their different perspective added to class discussions. Despite some of the difficulties such as later exam and homework returns, approximately 42% of students of both classes at UMass Dartmouth indicated in the exit survey that having the remote site actually helped the class while an additional 42% of them said that the addition of the remote site made no difference. It is impossible to make any statistically significant academic performance comparisons because of the limited sample sizes in these courses and the lack of a control group. However, the new courses seemed to result in higher success levels for each student. The instructor saw generally similar average performance on routine exam problems when compared to previous courses. This was true even though students who would have dropped in the past stayed and took the exams. In addition, students seemed to be better able to make progress on more difficult problems that were different from homework.
Acknowledgments This material is based upon work partially supported by the National Science Foundation under Grant Number DUE-
9551815. Support also came from the Davis Educational Foundation.
References 1.
Johnson, David W., Johnson, Roger T. and Smith Karl A., Active Learning: Cooperation in the College Classroom, Interaction Book Company, Edina MN, 1991. 2. Smith, Karl A., “Cooperative Learning: Effective Teamwork for Engineering Classrooms,” Proceedings of the Frontiers in Education Conference, November 1-4, 1995. 3. Pendergrass, N. A., “Using Computers, Simulators and Sound To Give Hands-on Experience”, Proceedings of the ASEE National Conference, June 23-26, 1996. 4. Wilson, Jack M., “The CUPLE Physics Studio,” Core Engineering, School of Engineering, Rensselaer Polytechnic Institute, 1995. 5. Pendergrass, N. A. and Sun, Gregory, “Using Computer Assisted Collaborative Learning For High Quality Distance Education,” Proceedings of the Frontiers In Education Conference, November 6-9, 1996.
