Remote Instruction in a Large Lecture Class

Electrical Engineering & Computer Science (Course 6)
Eric Grimson (courtesy The Tech: Sarang Kulkarni)
Eric Grimson
John Guttag
Fredo Durand
Ana Bell
Digital Innovations & Tools
Remote Teaching

6.0002 (Introduction to Data Analysis & Computational Experiments) is a half-term subject co-taught by professors Eric Grimson, John Guttag, and Fredo Durand, with logistical assistance from Dr. Ana Bell. Building on 6.0001, 6.0002 typically draws 400 students—90% undergraduates and 10% graduate students—with representation across almost every department at MIT.

Below, Grimson shares his insights and reflections on how they tackled the challenges of migrating the subject online.

Normal term: lectures, recitations, assessments, homework
Typically, 6.0002 comprises two lectures a week with optional recitations for diving into topics in more detail. Assessments consist of four micro-quizzes and weekly ‘finger exercises’ (short questions linked to the lectures, accessed via Residential MITx, and graded automatically). Five online problem sets are auto-graded with instant feedback to students using CAT-SOOP. When necessary, special cases are reviewed by staff. Normally, a TA or undergraduate Lab Assistant would also meet with each student for an additional “checkoff” session, in which students would describe and discuss their solutions.

Remote term: MITx, live & taped lectures, Zoom office hours
In the post-campus environment, 6.0002 relied extensively on MITx/edX resources. The MITx MOOC for 6.0002—one of the first ever created—provided a comprehensive array of materials, readings, self-paced problem sets, and lecture videos. In spite of the availability of a content-rich MOOC, Grimson and his co-instructors continued giving newly prepared live lectures on schedule. They intended to give students a sense of a continuity, even on Zoom. Lectures were delivered live, recorded, and then distributed for students in other time zones, to view asynchronously on their own schedules and preferred speed, or for review by students who wished to revisit elements of the lecture. Roughly half of the class attended the live lectures and the other half accessed the saved videos from the live lectures.

6.0002 office hours were handled online. When a student wished to interact with someone from the instructional team, they would post a request to a queuing system. Once they reached the head of the queue, a Zoom link would automatically be created, connecting the student to an LA or TA. Screen sharing allowed students to discuss their work in detail.

Although students were taking the class from all over the world, LAs and TAs were also dispersed globally. Because of this, it was possible to schedule office hours that accommodated both students and LAs and TAs according to global location.

Challenges: checkoff sessions, quizzes
The checkoff sessions were eliminated in the post-campus environment. However, subsequent remote iterations of the subject will reinstate the sessions as an integral part of the curriculum.

In the absence of live proctoring, remote quizzes and tests also presented a challenge. Grimson and his team opted for assessments that were ‘open book’ but ‘closed internet’. At the end of each test, students were asked to sign an honor code statement affirming they had acted appropriately. Grimson believes that there were a very small number of cases of cheating out of 400 students. In the future, Grimson and his colleagues may use an online proctoring system.

Successes: Zoom lectures, transcriptions, and chat
By conducting lectures over Zoom, instructors could use screen sharing to easily switch between PowerPoints and Python IDE for code runs. The Zoom transcription system was reliable and quickly available to students. There were very few complaints about access or quality.

The chat feature in Zoom was used extensively during lectures and transformed how the class was taught. Students used the chat to pose questions that a faculty member other than the lecturer staff would monitor and answer. If a particularly relevant or interesting topic was raised in chat, it was added to the live, spoken discussion. The chat function allowed for more activity and engagement than would have typically occurred in a live lecture.

The Zoom poll system served multiple functions. It was used to probe student comprehension of lecture topics, as well as for fun interludes and community building.

Anecdotal student feedback further indicated that 6.0002 lectures over Zoom are more interactive than had been possible in person.