A typical academic degree focused on software engineering has little practical relationship with the industry it is named for, other than the occasional placement or internship. Unlike other professions such as medicine, dentistry and veterinary sciences, candidates do not need to participate in significant professional practice to earn their degree. Indeed, if we consider a traditional academic software engineering student they probably have far more experience constructing shiny new 'green-field' systems, than maintaining the old 'brown-field' systems found in industry, and generating most professional work. Consequently, there is growing enthusiasm for work-based learning programmes that provide an opportunity for candidates to cement abstract academic theory in concrete personal experience. Work-based learning software engineering students earn their degree by combining theory with actual practice in a professional environment.
Nevertheless, the intangible outcomes for much of software engineering has led to an industry obsessed with confidentiality, driven by concerns of employees smuggling source code to competitors or regulators. This obsession potentially presents a barrier to work-based learning schemes as employers prevent outsiders, even close higher education partners, from observing the systems and the source code that learners are working on. Learners may have the opportunity for concrete personal experience, but educators are barred from observing any such experience.
However, confidentiality agreements may not necessarily present barriers to assessment, but instead provide an opportunity to assess comprehension and transferable skills by requiring abstract descriptions and reports. This is the converse to the problem in some programming courses, where students submit code without demonstrating that they understand it and can discuss it in terms of the concepts taught.
We explore some models for software engineering work-based learning programmes that have the potential to maintain confidentiality while assessing learners' comprehension and ability. We invite discussion and criticism from conference attendees of the presented models, and are interested in potential partners for future collaboration.