SIGCSE, Birds of a Feather (BOF) Session “eTextbooks”

My blogging of these events is getting later and later but another BOF session from Thursday night, hosted by Professor Cliff Schaffer from Virginia Tech. As always, my words may not quite match those of noted speakers.

Overall, a really thought-provoking panel – we all want to write but we ant to get all of the new features, without necessarily really knowing what the features available are or what the cost will be. The fear of wasting time is a constant spectre over the eBooks market. If I make it, I want it to be useful and feature-rich for some time.

What does the term eText even mean? Is it a type of text, the platform, the concept – some intersection? The virtues are obvious: portability, additional features like hypertext and search – but is this it? What are the educational benefits?

Cliff’s main idea was that, for our educational purposes, eBooks support interactivity and, hence assessment. There are projects like OpenDSA, a data structures and algorithms course in the creative commons. They’ve got content, texts, visualisations and assessment. Once a student has finished, they appear to be confident that they have understood the material.

Looking at Khan Academy it’s easy to focus on the videos, when the assessment exercises and awards system is an equally important part. But this is for Maths which is (notionally) easier to generate problem variations for and assess the result, to allow exercise in variations.

When students interact correctly with this progress determination activities, they answer the question, get told of their mark and given feedback and can then go again. Why do we mean by interactivity? (NF note, I’ll blog on this some more, later.)

There are a lot of solutions in this space, including algorithm simulation environment – how can we go beyond the textbook? Do we need to abandon the idea of the textbook as a closed container – does it make any sense any more?

The Open University in the UK has split their material between paper and electronic – electronic because of all the features and paper because students feel ripped off without a paper copy! The electronic materials have three levels of response to assessment-based interaction: firstly mark and just note where errors occurred, on the second pass, mark and suggest materials, on the third pass, if still under performing, direct student to read the material again. This is a bespoke system, producing Flash, but they hope to move to HTML5 at some stage.

Other tools mentioned included CTAT (Cognitive Tutoring Authoring Tools), AlgoViz and the amazing interactive textbook system written in Python, thinkcspy.appspot.com. If we’re going to have systems like Khan Academy, we need them to decomposable and re-usable but it would be nice if their grading system (badges) could work with us.

On the thinkcspy.appspot.com site, Brad and David’s book (Luther college), customised by Christine Alvarado, contains mid-term grades, log files and then end of term survey. CodeLens, visualisation was most correlated with results. However, outside of class time, students did not use most of interactive elements. The night before a test they flipped through the book. To learn this content, they have to change behaviour. Had assessment items already built in to drive knowledge boundary forward but students chose not to engage with the book.

Mark mentioned a new NSF project in October – building CS books for HS students to allow them to learn CS. Can’t use apprenticeship model because HS students don’t have time to mentor or be mentored because of an already full curriculum. The curse of outreach is that we have to take the time to produce and try to jam this into a heavily prescribed and full curriculum, to interest students in something – we need a mechanism that people will consult outside but it’s obvious that people won’t (according to the above).