Can removing letter grades actually encourage students to master computer science?
In 2010, Harvard's CS50 attempted a radical experiment: replacing traditional grades with a simple pass or fail system. While initially met with resistance, the move toward an ungraded format changed the classroom dynamic, impacting student enrollment, gender diversity, and even the time spent studying.
The transition toward a Satisfactory/Unnecessary (SAT/UNS) model in CS50 was driven by a desire to foster exploration. The original goal was to allow students without a computer science background to dive into complex programming concepts without the paralyzing fear of a low letter grade. However, the first attempt in 2010 faced significant institutional hurdles. Some academic concentrations refused to grant credit for SAT grades, and Harvard's general education program mandated that all courses be taken for letter grades.
A successful compromise arrived in 2013, allowing students to choose between a letter grade or the SAT/UNS option. By 2017, the SAT/UNS format became the default for the course. The impact was immediate and measurable. The proportion of students opting for the ungraded track surged from 9% to 31% in a single year. Notably, 13% of the 671 students enrolled specifically because of this flexibility, and the course saw a significant rise in female participation, reaching 44%—the highest level in 29 years.
Critics often worry that removing grades leads to reduced effort, but the data suggests the opposite. Students in the SAT/UNS track reported spending more time on course material each week than those in the letter-graded track. Furthermore, when adjusting for prior experience, performance levels between the two groups were nearly identical. The initiative suggests that as long as students are held to the same rigorous expectations and must complete all work to earn a SAT, the absence of a letter grade can serve as a powerful tool for academic engagement.
Source: Toward an Ungraded CS50 - SIGCSE 2021
