Post Midterm Viva Voces - CS70

Class Type: Discrete Mathematics and Probability Theory

After the CS70 midterm, students present the solution to a problem they struggled with to a TA. Students can demonstrate improved understanding of course content and use it as an opportunity to reflect on their performance.

Authors:

Presentations

Project Source

Table of Contents

  1. Introduction
  2. Course Context
  3. Learning Objectives
  4. Student Assignment
  5. Instructor Guids
  6. Requirements
  7. Mock Implementation

Introduction

Assessments shape both what students choose to learn and how they engage with the material. In computing education, the design of assessments therefore plays a central role in cultivating deep conceptual understanding, productive study habits, self-efficacy, a growth-oriented mindset, and collaborative problem-solving skills. Yet the prevailing assessment model in university-level computer science courses is that of high-stakes written midterm and final exams. These exams concentrate a semester’s worth of learning into a handful of timed, stressful events that may privilege performance under pressure over comprehension.

The rapid adoption of AI-assisted tools has altered the landscape of how university students study, complete homework, and prepare for assessments. As these tools continue to evolve, traditional written exams may no longer capture the full range of student learning or incentivize the behaviors that educators hope to encourage.

We’ve designed a post-midterm oral assessment for CS70, which we will refer to as a CS70 viva voce. After the midterm is released and returned, each student selects one long answer midterm problem to analyze in depth. During a one-on-one session with a TA, the student delivers a brief (seven to ten minutes) oral presentation on the midterm problem. Students have one index card for notes and whiteboard use is encouraged.

Course Context

  • A required lower-division course that introduces students to discrete mathematics and probability; it serves as introduction to theoretical CS.
  • Every semester, around 400 students enroll in the course.
  • The course includes two major written exams: a midterm and a cumulative final exam.
  • Historically, exams are challenging, resulting in average scores hovering around 50%.
  • Constrained by staff hours and workload: due to the large number of students, and the comparatively small course staff (29 paid staff members, as of Fall 2025).

Learning Objectives

The learning objectives for students in CS 70 are as follows.

  • Demonstrate their ability to analyse mathematical working and identify mistakes made.
  • Gain the ability to comprehend, reflect on, and analyze mathematical writing (CS70’s course notes and the midterm solution set).
  • Present mathematical reasoning with clarity and concision.
  • By preparing for conceptual follow-up questions, students will gain experience exploring technical concepts in-depth, enough that they are able to justify ideas from the course, and understand how they fit into the broader context of the course.
  • Learn to identify misconceptions, pitfalls, and other common mistakes.

Student Assignment

Students are asked to prepare an explanation that:

  • Introduces the key concepts related to the question from the notes (eg. formulas).
  • Presents the solution to the problem using these concepts introduced and justifies each step of work.
  • Explains the misconceptions in the student’s original solution, if the question was attempted.

They will deliver this explanation to a CS70 TA or tutor. If they’re working with a partner, they will be allocated up to ten minutes for their explanation. Otherwise, they will be allocated up to seven minutes for their explanation. During the explanation, the TA or tutor will ask them questions to clarify their explanation and their understanding of the selected question’s topic. To access the student facing handout, navigate from the Project Google Drive to “Student Handout.pdf.”

Instructor Guids

On the staff side, each staff member will prepare to administer oral exams for at least one midterm problem. In doing so, each staff member should brainstorm, collaboratively, a couple of conceptual questions regarding the midterm problem. The staff facing handout can be found from the Project Google Drive as “Staff Handout.pdf.” Additionally, staff members will grade students using a check-based rubric. This rubric can be found in the Project Google Drive as “Grading Sheet.pdf.”

Requirements

To run this assignment, the following resources need to be allocated:

  • Two hours of staff time (per staff member) for examination preperations.
  • 10 to 15 minutes of staff time (per student) to run the examination.
  • Adequate office hours and staff support on Ed.

Staff also need access to recording equipment and classrooms with whiteboards or blackboards.

Mock Implementation

We’re yet to implement this assignment. However, we have ran a small trial of the assignment. As part of the trial, we solicited feedback from both the staff and the students. For example:

image

The entire report on the experiment can be found here (“Final Report.pdf” in the Google Drive folder). The conclusions of our experiment were:

  • Students gain a lot from 1-1 interactions with teaching staff.
  • Viva voces were good opportunities for students to learn through teaching. Students felt like the assignment was meaningful.
  • Running the mock assignment was logistically challenging on its own; suggests that doing this at 100x the scale would be stressful.
  • Grading is required to make students do the assignment; however, subjectivity of grading is a concern. Would it be possible to design grading schemes that better service students?