How Harvard’s CS50 scales introductory computer science for students of varying experience

As introductory computer science classes expand, instructors face the challenge of balancing support for beginners with advanced tasks for experienced coders. This video explores how Harvard University’s CS50 course uses differentiated instruction and assignment design to accommodate diverse student comfort levels.

To manage growing variability in student experience, CS50 groups learners into sections based on their comfort levels. Students who report lower comfort receive additional support through smaller section sizes. The course also features problem sets with tiered floors and ceilings, ensuring that assignments remain appropriately challenging for every student regardless of their background.

The team currently uses score normalization to assess students across these different comfort levels. Future initiatives aim to provide deeper support for the least comfortable students, particularly those whose secondary education may not have fully prepared them for university-level computer science.

Source: SIGCSE 2023 - Differentiating for Comfort with Computer Science