Rethinking assessment in early computing courses

This research-to-practice paper considers the impact of theory and findings from cognitive science research on assessments in computing courses. While the pandemic-forced virtual learning interrupted many standard classroom routines, these interruptions also offered opportunities to innovate beyond common pedagogical tropes. One educational mainstay, the proctored test, became particularly difficult to administer in virtual environments. Ensuring that a test only reflects an individual’s learning without external support became more cumbersome than the alternatives – alternatives that might offer better long-term options for assessing learning.Many educators question the equity and effectiveness of high-stakes testing, with assessments of computing skills introducing even more challenges. Tests typically require students to demonstrate learning apart from computers, assuming such knowledge transfers equally to pen and paper. Advances in cognition suggest that learning may not transfer between activities in a timely and effective manner, particularly when much of the work is hands-on. Paper tests may not accurately measure a curriculum’s learning objects giving students inaccurate feedback on their abilities. Non-traditional students with professional experience in coding may be disenfranchised as their practical knowledge fails to help on tests. Some students with otherwise strong GPAs may become frustrated when unable to tackle open-ended problem-solving tasks in later courses. Both groups are learning but need assessments, policies, and grading approaches that measure all types of learning.This paper describes integrating theory and research from the cognitive sciences with practice in the computing classroom. The foundation for the discussion is based on a new theory of programming cognition called TAMP. TAMP suggests that expert programmers gather and apply knowledge both explicitly and implicitly. Traditional testing may not investigate all types of knowledge and thus may miss important aspects of a new programmer’s development. Theoretical advances drive new research and can have a vital impact on classroom practices. This discussion reflects on how theory (and the pandemic) drove changes to assessment in computing classrooms. The hope is that discussing how theory and research drive practice can inform future research and inspire additional practical improvements.