Online Learning: Technologically Enhanced Education
in Electrical Engineering and Computer Science

Dates

July 2000 — June 2003

Principal Investigators

Professor Tómas Lozano-Pérez (Department of Electrical Engineering and Computer Science)
Professor Eric Grimson (Department of Electrical Engineering and Computer Science)

Goal

The goal of Online Learning is to boost comprehension of material in introductory computer science by enabling students to learn at their own pace and schedule, obtain immediate feedback on their progress, and enable instructors to spend more time developing new content and interacting with students.

Overview

The Electrical Engineering and Computer Science (EECS) faculty has implemented major changes in the format and content of several key courses. They are developing and using online interactive technologies that enable students to learn at their own pace and on their own schedules, and to get immediate and complete feedback on their mastery of material. This also provides the faculty more time to develop new content, and allows teaching assistants to spend their time in face-to-face interactions with the students, rather than spending all their time grading.

This work uses different combinations of online presentations based on narrated PowerPoint slides and Web-based assignments that provide immediate feedback. These assignments include programming and circuit design, and provide students with a wide range of challenges.

These technologies define a new genre of course materials that goes far beyond printed textbooks or web sites. This research has been used in three of the large undergraduate Computer Science subjects in EECS (6.001, 6.034 and 6.004). Extensive surveys have been conducted in both 6.001 and 6.034, confirming that student reaction has been quite positive.

More Project Details from the Principal Investigators

We believe that the interactive delivery system will dramatically impact the education of one-third to one-half of the entire undergraduate population. Experience shows that attendance in large lectures drops off as the term progresses, despite the presence of excellent lecturers. This often causes students to miss an important opportunity to experience the material and methods of a subject. Similarly, effective feedback from problem sets in large classes is minimal because staffing issues often imply that student work is graded and returned several weeks after the student has completed it, with the result that the students have often forgotten what they did by the time they get the graded work, and hence miss an important educable moment. Pilot studies in 6.001 and 6.034 suggest that the online system yields much higher participation by students, often on their schedule, rather than the institute's schedule, and immediate feedback can reduce wasted effort and improve understanding by the students. We hope that a detailed evaluation of the methods will demonstrate the improved effectiveness of the method in helping students learn the material, and if so, that this methodology could serve as a wide-spread tool for many courses at MIT.

Our intent is to build an infrastructure for developing course-tuned interactive delivery systems that is platform independent, is applicable to a wide range of course styles, and has an easily used interface so that instructors can instantiate and maintain their own tutors. As a consequence, it would be possible to distribute the infrastructure, together with guidance on how to use it, beyond MIT. As the project progresses, we will explore alternative plans for distributing the infrastructure, such as through OpenCourseWare, through a distribution with Microsoft or other software suppliers, or as open source distribution.

Conclusion by the Principal Investigators

• We have shown that it is possible to use online presentations to deliver high-quality educational content in a way which the students feel is effective, even over the traditional presentation modes (ie., lectures, recitations). The students value schedule flexibility and the ability to go back and study the lectures/recitations.
• We have shown that online assignments that go beyond multiple choice (including programming and circuit design) have great advantages for the students and the staff. They provide students with immediate feedback as to what they do and do not understand. They also cut down on routine grading.
• Together, these two components constitute the "textbooks of the future." People have been building books on CD-ROM for a while but these Web-based systems have a number of advantages. They can be done anywhere, are relatively cheap to develop and upgrade (on the fly) and provide useful reporting as to the progress and use of the materials by the students.

The 6.004 lab submission system has worked well. The lab staff can spend their time helping students debug their designs rather than doing check offs. The students don't have to wait in long queues in the lab in order to complete the assignment. A less obvious side-effect is that the verification procedure often is very effective at pin-pointing where a design has failed, usually leading to a quick fix by the student. This is a direct result of the automated exhaustive cycle-by-cycle evaluation of the circuit response, where before we used much simpler checks.

"I really liked 6.001. I like the online lectures because I actually make sure I understand them well, whereas in live lectures, if I didn't understand something, I just let it go. I learn most of the stuff on my own anyway, and recitations and tutorials reinforce everything better. That's how it is for most classes—I just blindly copy down what is written on the board and learn everything through books. 6.001 is the first class at MIT where I feel like I learned the material well. I think it is a great class!" — Student in EECS 6.001

Project Output

Presentations

Experience with Web-Based Computer Science Education, Leslie Pack Kaelbling, Consortium for Computing in Small Colleges, Seventh Northeaster Conference, April 2002.

On-line On-campus: On-line Presentations and Interactive Problems in Large Introductory CS Courses, Tomas Lozano-Perez, CMI Workshop, Cambridge England, June 17 2002.

Links

Xtutor Outreach