COOKBOOK INTERACTIONS Cookbooks, both in print and ebook form, have specific characteristics. They typically include, beyond the recipes themselves, supporting fea- tures such as background information on the recipes, general cooking instructions, glossaries, and charts for measurement conversions. Eat St. followers revealed through our survey that they are already buying cookbooks for their devices, with a preference for cookbooks that are curated and organized around themes, and which include interactive features such as “cooking modes” for following recipes from the tablet screen while cooking. The Eat St. Cookbook is articulated through an interactive map that allows users to access recipes based on location, while a secondary con- tents table allows users to search for vendors or recipe titles. Each recipe has distinct sections: a vendor section, with information about the food cart owner's history; facts about the recipe; images of other dishes by the same vendor; a clip from the episode featuring the cart; and links to follow vendors on Twitter or Facebook. The recipe section allows users to see an overview of ingredient lists and cooking instructions, while also being able to access a large-scale cooking mode, which shows each individual step on the screen. Users can also share the recipe by email, post it to Facebook, or share it on Twitter; these social media and shar- ing features are part of the standard expectations of our users. PAGES AND SCROLLS As noted in early user-centred studies of ebook design, adherence to the paper book metaphor has a significant impact in the usability of ebooks. [4] The parsing of content into pages, as opposed to continuous scroll- ing, helps give readers a “sense of place” within the ebook and marks their progress in the book or section (Figure 2). Whereas the notion of pages will be contestable for some kinds of content, in the case of the Eat St. Cookbook, pages are a crucial feature that not only parse content into manageable bits, but also aid the reader in generating a mental map of the kinds of content included for each recipe stack: page 1 of the stack introduces the recipe with a full image, page 2 features the food truck story, page 3 includes the full recipe with ingredients and instructions. An unobtrusive marker shows the location of each page in the stack. USER TESTING After several iterations of the ebook were tested rapidly, informally and early in the process by team members, we refined the concept and built an extensive prototype that included all navigation features and a selection of ten recipes with full interactivity in place. In order to asses user experience and gather qualitative insights from users’ responses to the ebook narrative, we conducted a formal user testing trial, approved by the University’s Research Ethics Board, with 10 representative-user participants. Participants were asked to perform representative tasks that ranged from highly specific, directed but less specific, to open-ended explora- tions of the ebook. The trial was organized into four sections: the first asked introductory questions about participants’ familiarity with tablets and their cooking habits; the second asked participants to freely explore the ebook for five minutes to familiarize themselves with it, during which time they were encouraged use the think aloud technique to ver- balize what they were doing and thinking; the third asked participants to perform specific tasks consisting of reaching certain points in the cookbook using different navigation paths; and the final section asked a number of specific questions about the cookbook design, its interactive features, and participants’ overall perception of the ebook. Our findings confirmed that the overall experience and navigation of the cookbook was intuitive, coherent and pleasurable, though some problem areas were identified for further refinement. These included: + Multiple screen swipes required to access the table of contents or recipe map, which is consistent with literature on iPad usability. [3] “Back to Contents” buttons were implemented throughout to resolve this issue. « Ambiguity of icons and text, and their possible performance as but- tons. Native DPS buttons don’t provide direct visual feedback upon touch as do HTML based buttons. + — Use of image sequences (slides). Users were uncertain of whether to swipe or tap to move from image to image. We resolved this issue by setting both swipe and tap as options for the sliders. + Small images as buttons for activating a larger view of the image were ignored. We included further visual and textual cues to make their function more apparent. « Arrows indicating “previous” and “next” steps in sliders were considered buttons by all users, even though they weren't set up as such. All arrows were converted to buttons. + The use of DPS’s native scrollable frame for the table of contents was very ineffective and frustrating for users. We reconfigured the contents to multiple pages. * Participants were unsure of whether to access recipes by tapping directly on the map or on the label that identified each region. Where multiple access points were available, all were converted to links. CONCLUSION The final Eat St. ebook incorporated revisions based on insights gained through user testing and was fully produced with 50 recipes from season three of the program. The overall narrative and visual structure presented a curated experience of street food as a culinary trip and addressed expectations of users both in terms of the ability to share content as well as interactive features that facilitate cooking from a tab- let. A workflow for content generation and curation as well as produc tion was included as part of the deliverables for the project. As a result, we provided Invoke Media with not only a ready-to-commercialize product, but also a workflow and template for future collections of rec ipes that can be extended to multiple platforms, allowing our industry partner to benefit from Emily Carr’s expertise in ebook design. REFERENCES [1] Blackberry, Apple, Kobo top brands in Canada’s mobile device market. fpsos North America (2011). Press Release. http://www.ipsos-na.com/news-polls/pressrelease. aspx?id=5399 [2] Conneally, T. Canadians prefer Blackberry, Kobo — Americans favor Android, Kindle. Beta News (2012). http://betanews.com/2012/04/20/canadians-pre- fer-blackberry-kobo-americans-favor-android-kindle/. [3] Nielsen, J. [Pad usability: Year One. Nielsen Norman Group (2011). http://www.nngroup.com/articles/ipad-us- ability-year-one/ [4] Wilson, R., Landoni, M., and Gibb, F. A user-centred approach to e-book design. The Electronic Library 20, 4 (2002), 322-330. INTERACTIVITY we =