Spatially-stable menus, FastTap, use grids to display commands on touch tablets, where grid lines serve as landmarks, enabling users to quickly learn command locations. While these menus can expand the grid to allow more commands, repetitive lines can weaken landmark aid and hinder spatial learning. While artificial landmarks showed spatial learning benefits in desktop menus, little is known if landmarks can enable spatial learning on tablets with many commands. So, we conducted two studies with different landmarks (standard grid, menu-backdrop image, and coloured blocks) in single-tab and multi-tab FastTap menus with large command sets. Results indicated that people developed spatial memory of commands in all interfaces across two studies. While coloured blocks substantially improved command selection performance in single-tab, both landmarks notably improved selection speed at least during the early learning stages in multi-tab. People also issued significantly more expert selections with landmarks in multi-tab. We demonstrate that artificial landmarks can aid spatial learning in tablets with large command sets, which will help the future design of tablets.

Graphical User Interfaces present commands at particular locations, arranged in menus, toolbars, and ribbons. One hallmark of expertise with a GUI is that experts know the locations of commonly-used commands, such that they can find them quickly and without searching. Although GUIs have been studied for many years, however, there is still little known about how this spatial location memory develops, or how designers can make interfaces more memorable. One of the main ways that people remember locations in the real world is landmarks – so we carried out a study to investigate how users remember commands and navigate in four common applications (Word, Facebook, Reader, and Photoshop). Our study revealed that people strongly rely on landmarks that are readily available in the interface (e.g., layout, corners, and edges) to orient themselves and remember commands. We provide new evidence that landmarks can aid spatial memory and expertise development with an interface, and guidelines for designers to improve the memorability of future GUIs.

Learnability is important in graphical interfaces because it supports the user's transition to expertise. One aspect of GUI learnability is the degree to which the icons in toolbars and ribbons are identifiable and memorable - but current "flat" and "subtle" designs that promote strong visual consistency could hinder learning by reducing visual distinctiveness within a set of icons. There is little known, however, about the effects of visual distinctiveness of icons on selection performance and memorability. To address this gap, we carried out two studies using several icon sets with different degrees of visual distinctiveness, and compared how quickly people could learn and retrieve the icons. Our first study found no evidence that increasing colour or shape distinctiveness improved learning, but found that icons with concrete imagery were easier to learn. Our second study found similar results: there was no effect of increasing either colour or shape distinctiveness, but there was again a clear improvement for icons with recognizable imagery. Our results show that visual characteristics appear to affect UI learnability much less than the meaning of the icons' representations.

Time-lapse videos are often navigated by scrubbing with a slider. When networks are slow or images are large, however, even thumbnail versions load so slowly that scrubbing is limited to the start of the video. We developed a frame-loading technique called spread-loading that enables scrubbing regardless of delivery rate. Spread-loading orders frame delivery to maximize coverage of the entire sequence; this provides a temporal overview of the entire video that can be fully navigated at any time during delivery. The overview initially has a coarse temporal resolution, becoming finer-grained with each new frame. We compared spread-loading with traditional linear loading in a study where participants were asked to find specific episodes in a long time-lapse sequence, using three views with increasing levels of detail. Results show that participants found target episodes significantly and substantially faster with spread-loading, regardless of whether they could click to change the load point. Users rated spread-loading as requiring less effort, and strongly preferred the new technique.

Document readers with linear navigation controls do not work well when users need to navigate to previously-visited locations, particularly when documents are long. Existing solutions - bookmarks, search, history, and read wear - are valuable but limited in terms of effort, clutter, and interpretability. In this paper, we investigate artificial landmarks as a way to improve support for revisitation in long documents - inspired by visual augmentations seen in physical books such as coloring on page edges or indents cut into pages. We developed several artificial-landmark visualizations that can represent locations even in documents that are many hundreds of pages long, and tested them in studies where participants visited multiple locations in long documents. Results show that providing two columns of landmark icons led to significantly better performance and user preference. Artificial landmarks provide a new mechanism to build spatial memory of long documents - and can be used either alone or with existing techniques like bookmarks, read wear, and search.

Navigating to previously-visited pages is a trivial yet fundamental task in linear control-based document viewers. These widgets e.g., scrollbars often do not work well particularly for long documents. Existing solutions try to tackle this issue with bookmarks, search, history, and read wear but limited in terms of effort, clutter, and interpretability. To improve the revisitation support in long documents, we investigated the use of artificial landmarks similar to the visual augmentations available in physical books: coloring on page edges or indents cut into pages. We developed several artificial-landmark visualizations to represent page-locations in the scrollbar for many hundreds of pages long documents, and tested them in studies where participants visited multiple locations in long documents. Results indicate that using two columns of landmark icons significantly improved revisitation performance and preferred by users. Our two-level artificial-landmark augmented scrollbars can be a new way to support spatial memory development of long documents - and can be used either in isolation or in congregation with current techniques.

The capacity of spatial multi-touch menus such as FastTap is limited by device screen size. We explore the idea of using multiple tabs to increase capacity - multiplexing the tablet's screen space so each location holds multiple items. Earlier work has shown potential of this idea for smartwatches, but no evaluations have considered larger devices. To assess issues with interference-based errors and spatial memory development, we built two FastTap systems with multiple tabs and conducted two studies. We first tested user learning of 16 targets with a training game, and found that participants easily adapted to the multi-tab model, were able to perform memory-based shortcuts, and made few interference-based errors. The second study used realistic drawing tasks and showed that people successfully used the multi-tab FastTap system, with 88% of selections made using shortcuts by the study's end. Our work demonstrates that spatial memory can successfully be multiplexed, and that tabs are a promising way to increase command set sizes for spatial interfaces.

Spatial memory is a powerful way for users to become expert with an interface, because remembering item locations means that users do not have to carry out slow visual search. Spatial learning in the real world benefits greatly from landmarks in the environment, but user interfaces often provide very few visual landmarks. In this paper we explore the use of artificial landmarks as a way to improve people's spatial memory in spatially-stable grid menus called CommandMaps. We carried out three studies to test the effects of three types of artificial landmarks (standard grid, simple anchor marks, and a transparent image) on spatial learning. We found that for small grid menus, the artificial landmarks had little impact on performance, whereas for medium and large grids, the simple anchor marks significantly improved performance. The simple visual anchors were faster and less error-prone than the visually richer transparent image. Our studies show that artificial landmarks can be a valuable addition to spatial interfaces.

Command selection on large multi-touch surfaces can be difficult, because the large surface means that there are few landmarks to help users build up familiarity with controls. However, people's hands and fingers are landmarks that are always present when interacting with a touch display. To explore the use of hands as landmarks, we designed two hand-centric techniques for multi-touch displays -- one allowing 42 commands, and one allowing 160 -- and tested them in an empirical comparison against standard tab widgets. We found that the small version (HandMark-Fingers) was significantly faster at all stages of use, and that the large version (HandMark-Multi) was slower at the start but equivalent to tabs after people gained experience with the technique. There was no difference in error rates, and participants strongly preferred both of the HandMark menus over tabs. We demonstrate that people's intimate knowledge of their hands can be the basis for fast and feasible interaction techniques that can improve the performance and usability of interactive tables and other multi-touch systems.

Fast command selection is important for touch devices, but there are few techniques that allow accelerated selection while still providing a large command set. HandMark menus [25] propose the use of the hands as landmarks for fast memory-based selection. However, the original HandMark menus rely on bimanual operation, and earlier studies provided only limited evidence for the value of hand-based landmarks as a reference frame for spatial memory. In this paper we address these limitations. We introduce adapted HandMark menus that can be operated with one hand (while the other hand holds the tablet); the new version changes bimanual selection operations into sequential actions with one hand. We carried out three studies of these HandMark menus. The first study showed that the adapted menus still allowed fast performance and the development of spatial memory, even with one-handed use. The second study focused on the value of hands as landmarks, by comparing HandMark menus against a hidden popup menu. This study showed that using the hand as a reference frame significantly improved performance, and was strongly preferred by participants. Our work extends HandMark menus and shows that they are an effective selection method for tablets, and provides new evidence about the value of the hands as a spatial landmark for interaction.

Although rapid command selection is important for touch devices, only a few techniques allow accelerated selection while still providing a large command set. HandMark menus [3] propose the use of the hands as landmarks for fast memory-based selection. However, the original HandMark menus for tables rely on bimanual operation - not suitable for tablets, and earlier studies provided only limited evidence for the value of hand-based landmarks as a reference frame for spatial memory. We introduce adapted HandMark menus that can be operated with one hand (while the other hand holds the tablet); the new version changes bimanual selection operations into sequential actions with one hand. Three studies of these HandMark menus showed that using the hand as a landmark significantly improved performance and allowed fast spatial memory development, even with one-handed use. Our work extends HandMark menus and shows that they are an effective selection method for tablets, and provides new evidence about the value of the hands as a spatial landmark for interaction.

Code clone visualizations (CCVs) are graphical representations of clone detection results provided by various state-of-the-art command line and graphical analysis tools. In order to properly analyze and manipulate code clones within a target system, these visualizations must be easily and efficiently comprehensible. We conducted an eye-tracking study with 20 participants (expert, intermediate, and novice) to assess how well people can comprehend visualizations such as Scatter plots, Treemaps, and Hierarchical Dependency Graphs provided by VisCad, a recent clone visualization tool. The goals of the study were to find out what elements of the visualizations (e.g., colors, shapes, object positions) are most important for comprehension, and to identify common usage patterns for different groups. Our results help us understand how developers with different levels of expertise explore and navigate through the visualizations while performing specific tasks. Distinctive patterns of eye movements for different visualizations were found depending on the expertise of the participants. Color, shape and position information were found to play vital roles in comprehension of CCVs. Our results provide recommendations that can improve the implementation of visualization techniques in VisCad and other clone visualization systems.