Management, structures and tools to scale up personal advising in large programming courses

Abstract

We see programming in higher education as a craft that benefits from a direct contact, support and feedback from people who already master it. We have used a method called Extreme Apprenticeship (XA) to support our CS1 education. XA is based on a set of values that emphasize actual programming along with current best practices, coupled tightly with continuous feedback between the advisor and the student. As such, XA means one-on-one advising which requires resources. However, we have not used abundant resources even when scaling up the XA model. Our experiments show that even in relatively large courses (n = 192 and 147), intensive personal advising in CS1 does not necessarily lead to more expensive course organization, even though the number of advisor-evaluated student exercises in a course grew from 252 to 17420. A thorough comparison of learning results and organizational costs between our traditional lecture/exercise-based course model and XA-based course model is presented.