A CONTINUAL IMPROVEMENT PARADIGM FOR MODERN ONLINE TEXTBOOKS

Traditional hardcopy textbook publishing follows an "editions" paradigm, wherein a new edition of a textbook is developed and released every few years. The edition paradigm is necessary in part due to the time and cost of producing and printing a new edition. The evolution of online textbooks over the past decade has largely followed that same editions paradigm, with an online textbook version mirroring a hardcopy textbook version of the same edition, either exactly or in compacted or expanded form.

However, some recent new online textbooks are written exclusively for the web, having no hardcopy textbook version. Such an online-only textbook is freed from the time and cost of producing and printing a new edition. As such, some such textbooks now follow a "continual improvement" paradigm, wherein the textbook is updated frequently, akin to how web applications (like a web browser) are updated, with such updates often being entirely transparent to the user.

This paper describes our experiences with a continual-improvement paradigm for textbooks used by over 60,000 students in the past two years, including several discovered advantages, and numerous challenges.

Several advantages have been encountered. Feedback from users can be supported via feedback buttons embedded throughout the textbook. With appropriate processes, reported errors can be fixed and pushed to all users almost immediately, resulting in material that quickly converges into having few or no significant errors. Errata pages are thus no longer necessary. Furthermore, usage data as well as user feedback can guide semi-annual improvement passes by authors over the textbook content. For example, frequently mis-answered embedded learning questions can inform authors as to what content is in most need of improvement. Another advantage is that the subscription model for online textbooks eliminates the need to artificially create new editions every few years to prevent excessive revenues from being lost to the used-book/rental market. Instead, improvements can be made on a more natural basis, with sections or chapters solely based on content need, and not as a defensive market strategy. Also, as new related textbooks are published, existing textbooks can immediately benefit via mixed-content offerings.

A key challenge relates to content stability. Instructors may come to expect the presence of particular content, like a particular figure, and thus significant modifications of material must be done in a manner that doesn't disorient instructors. With more frequent updates, another challenge is informing instructors of the changes, without inundating instructors.

Beyond the above-experienced advantages and challenges, online textbooks enable additional possibilities, such as instructors improving content by adding custom content, instructors contributing content (such as learning questions) that is then used by others, users rating such content such that the best content becomes recognized and perhaps becomes part of the core content, analyzing how students use content to determine which content appears to be most useful, and much more.