At the Helsinki University of Technology, instructors facing larger and larger introductory class sizes have moved beyond simply having Computer-Aided Assessment into the realm of fully Automated Assessment (AA) [18]. Malmi, Korhonen, and Saikkonen acknowledge that fully automated assessment does not give students the depth of feedback that could be given if resources were available for more human graders and smaller class sizes, but at the point that courses have become too large to handle any other options, AA is far superior to the other obvious alternative: self assessment or voluntary programming assignments.
AA has some benefits, some of which are appealing even when compared to traditional grading. The most obvious benefit is the reduction in resource requirements, as mentioned. Secondly, the system can give back feedback at any time of day or night. This second property leads naturally into the third: since feedback is returned very quickly and entirely objectively, it is sensible to allow students to resubmit based on that feedback. This is in line with the constructivist view of pedagogy that has taken hold in the educational community. If the constructivist view is correct, then there is certainly some value in AA. Whether or not there is value in human-checked assessement using the constructivist resubmission paradigm is not addressed in this work.