Bypassing Lists: Accelerating Screen-Reader Fact-Finding with Guided Tours
Tao Yang, Prathik Gadde, Robert Morse, Davide Bolchini · 2013 · Proceedings of the 15th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS) · doi:10.1145/2513383.2513435
Summary
This paper investigates whether guided tours — a navigation pattern that linearly links content pages with next/previous controls — can help screen-reader users perform fact-finding tasks more efficiently than traditional index-based navigation, particularly when the index lacks adequate information scent (cues that help users select the right link). When an index on a website like BestBuy.com does not show the specific detail a user needs (e.g., number of USB ports), screen-reader users must repeatedly navigate from the index to content pages and back, re-gaining orientation on the index each time. This back-and-forth is especially costly in aural navigation because users must listen through list items sequentially. The researchers conducted a within-subject controlled study with 11 blind participants (ages 15-19) at the Indiana School for the Blind and Visually Impaired, comparing three navigation patterns — index, guided tour, and mixed pattern (index + guided tour combined) — across fact-finding tasks in 30-item collections. Tasks were designed so that index scent did not match the task cue, forcing users to explore content pages. The study varied whether the target item was in the first half (items 1-15) or second half (items 16-30) of the collection.
Key findings
Guided tours made screen-reader users up to 44% faster than index-only navigation when the target was in the first half of the collection: time-on-task dropped from 325.9 seconds (index) to 192.8 seconds (guided tour). Pages visited decreased by 41% (from 18 to 10.6). When the target was in the second half of the collection, guided tour still significantly reduced cognitive effort (1.43 vs. 2.20 for index, p<.017) and pages visited, though the mixed pattern proved most efficient for time-on-task. The mixed pattern combined the best of both approaches: it allowed users to jump into the collection from any index item and then navigate linearly, significantly reducing keystrokes in both collection halves. Participants overwhelmingly preferred guided tour and mixed pattern over index — 90% of participants who started with index in the mixed pattern immediately switched to guided tour navigation. Qualitative feedback was striking: one participant said index navigation "needs a lot of patience" while another noted guided tours provided clear position indicators ("like four out of ten") and eliminated the painful back-and-forth. Importantly, 50% of participants reported difficulties with index navigation while only 30% had problems with guided tour and 10% with mixed pattern.
Relevance
This research challenges a fundamental assumption in web design — that indexes are the best navigation structure for fact-finding tasks. While this assumption holds for sighted users who can visually scan lists and track visited links, it breaks down for screen-reader users who must linearly process list items and cannot easily maintain orientation when navigating back and forth. For web developers and accessibility practitioners, the practical implication is clear: supplement index pages with guided-tour navigation (next/previous links between content items) as a complementary access structure. This is especially valuable for e-commerce product listings, encyclopedic collections, search results, and any collection where index labels may not contain sufficient detail for all possible search criteria. The finding that the mixed pattern (offering both index and guided tour) performed best overall suggests that providing multiple navigation strategies — rather than choosing one — best serves diverse user needs.
Tags: screen reader · web navigation · blindness · visual impairment · information scent · guided tour · web accessibility · fact-finding · usability