An analysis of practitioner studies concludes that a gap exists between industry expectations of IT graduates and formal education, in particular in non-technical skills and knowledge. This paper reports on the final cycle of an Action Research project to examine and implement alternative learning environments for Software Engineering. A model based on reflective practice, founded on the evaluation of previous cycles applying Cognitive Apprenticeship and Problem-based Learning, was developed and implemented. This study looked at the alignment between student approaches to learning and the environment developed, in particular student disposition towards deep or surface learning. Although evaluation shows the student cohort achieved significantly higher scores in their assessment that those of previous offerings, it was notable that students who reported themselves as adopting surface approaches were less comfortable with the environment. However, they still exhibited deep learning characteristics when observed in a subsequent course.

While an understanding of student learning is fundamental in developing learning environments, alignment between discipline and learning is also critical in educating competent practitioners. Results of this study show that students placed in an environment that enables them to model professional practice, and reflect on that modelling, should be much better prepared for the workplace.