While Greenstone is a versatile tool for creating and managing digital collections, its interface is not well suited for senior users. As we age, our usability requirements for software change; the mismatch between interfaces designed for (and by) younger members of society and the interface usability requirements of the elderly can make software less usable for the elderly. This paper reports on an intensive usability study of the Greenstone ‘Librarian’ interface (supporting collection development and maintenance) with senior participants. The findings of this study contribute to a re-design of Greenstone for senior users.
The Greenstone digital library suite is intended for use by a broad swathe of adult users; the objective is that any individual can, with minimal training, organize documents and serve them over the Web (Witten et al, 2010). Research into interface and interaction design requirements for the elderly, however, suggests that Greenstone would be more usable and accessible to seniors if a special ‘senior mode’ system were available (Hawthorn, 1998, Hawthorn, 2000, Xie, 2003). As we age, we experience cognitive, visual, perceptual, and motor changes that make use of standard interfaces more difficult—and these difficulties are exacerbated by poor design (Xie, 2003). This paper reports on one phase of a re-design effort for the Greenstone ‘Librarian’ (which supports collection creation and maintenance, with the goal of making Greenstone more suitable for use by seniors to manage their personal document collections.
We informed our re-design through a literature review of interfaces and aging and two user studies; we report here on the second user study, which examines the usability of the Librarian by seniors (aged over 65) and identifies modifications implemented in a ‘senior mode’ prototype.
USER STUDY: DOCUMENT TYPES AND METADATA FOR A PERSONAL LIFE HISTORY COLLECTION
An intensive usability study was conducted over the standard Greenstone Librarian sub-system, to confirm that a re-design for seniors was necessary and to identify portions of the Librarian that must be re-worked. Eight seniors were recruited as participants (seven males, one female), all over 65 and with an average age of 74. They averaged 23 years of computing experience.
The participants were asked to follow through a tutorial on using the Greenstone Librarian. The tutorial was broken into three sessions (Exploring the structure of an existing collection; Creating a collection and adding metadata to its documents; Customizing a collection's search and browsing features), with each session lasting at least 90 minutes. The sessions were video-recorded. After each session, the participant was debriefed with a discussion of the Greenstone Librarian and the aspects of the tutorial tasks that they felt were difficult / easy, the problems that they experienced in completing the tutorial, and so forth. Participant performance on selected tasks within the sessions was timed.
A qualitative analysis of the video-recordings and post-session debriefings identified several problems with the Librarian common to most or all of the participants:
Confusing terminology: All eight of the participants had difficulty interpreting multiple labels in the Librarian interface. Some of the problematic terminology related to technical DL/IR terminology (e.g., stemming, case folding), some to computing terminology (e.g., related to file sizes), and some was idiosyncratic usage specific to Greenstone (e.g., ‘Enrich Document’). Where possible, these labels were modified to correspond more closely to normal usage (e.g., ‘Enrich Document’ was changed to ‘Describe Document’). Many of the technical labels were discovered to apply to concepts that were unlikely to be relevant to the needs of a senior constructing a personal life history digital library; the associated features were moved to a separate, ‘Advanced’ options tab.
Complex workflow: Developing a new collection in the standard Librarian interface requires the user to work both from left to right and from top to bottom, sometimes within a single screen and sometimes from one screen (tab) to another. On completing subtasks participants frequently commented on confusion as to what to do next, and this confusion frequently led to errors. The workflow in the ‘senior mode’ was modified to move primarily from left to right, and made consistent across all screens.
Egregious use of icons and popups: The standard Librarian includes both icons and text labels for many options, and uses tooltips extensively. The seniors frequently confused icons with tooltips, and vice-versa; they also were distracted by what they saw to be a random display of tooltips. The presence of both labels and icons was confusing as well; was this one option with two identifiers, or two separate options? Should the user click on an icon, the label, neither, both? Removing tooltips and nearly all the icons from the ‘senior mode’ Librarian interface, leaving it primarily text based, solved this problem.
Interface clutter: Extensive use of icons and inclusion of a multiplicity of options in the standard Librarian interface leaves it with a cluttered feel—font sizes are small and interface elements are closely packed. The clutter problem was partly solved as above, by moving infrequently required options to a separate tab and eliminating tooltips and icons. A minimum distance of 5 pixels between objects within a pane was enforced, to make it easier to distinguish between elements and reduce ‘overshooting’ when the user selects an object.
Interestingly, the interface guidelines for software for senior users (Hawthorn, 1998; Hawthorn, 2000) recommendation against the use of the color green as a background was not upheld for the participants. A red-based version of Greenstone was trialed with the participants, and their performance in distinguishing detail and completing tasks was not significantly affected by the background color—so the senior-friendly version of Greenstone remains green.
EFFECTS OF LEARNING AND PRACTICE
The above insights into difficulties with the Librarian's usability were primarily derived from qualitative, informal analysis of session videos and participant debriefings. These informal insights were indirectly confirmed by an analysis of the consecutive times taken by each participant to go through the set of steps to create a digital collection.
Each of the three study sessions required each participant to create at least one digital collection; in total, each participant went through the creation process at least nine times (Figure 1; timings for participants to create collections from previously identified and tagged documents). Given this amount of practice, it should be expected that the participants would ‘proceduralize’ the steps involved—and that their individual collection creation times should decrease with practice. No such significant decrease was observed (paired t-test of initial and final collection creation times resulting in a p-value of 0.113 over a 95% confidence interval). Clearly all the participants were finding it difficult to both use and to become skilled with the standard Librarian sub-system, despite significant training and practice (three sessions, with a minimum of 90 minutes per session).
The Librarian sub-system of Greenstone is clearly shown to have usability hurdles for senior users. The usability issues identified in this study suggested modifications to the standard Greenstone Librarian, which have been implemented in a prototype. Further evaluation of the ‘senior mode’ prototype is necessary to ensure that these modifications address the problems—though it is heartening to note that these changes appear likely to benefit a broad range of users, beyond the elderly.