Faster Document Navigation with Space-Filling Thumbnails

Scrolling is the standard way to navigate through many types of digital documents. However, moving more than a few pages can be slow because all scrolling techniques constrain visual search to only a small document region. To improve document navigation, we developed Space-Filling Thumbnails (SFT), an overview display that eliminates most scrolling. SFT provides two views: a standard page view for reading, and a thumbnail view that shows all pages. We tested SFT in three experiments that involved finding pages in documents. The first study (n=13) compared seven current scrolling techniques, and showed that SFT is significantly faster than the other methods. The second and third studies (n=32 and n=14) were detailed comparisons of SFT with thumbnail-enhanced scrollbars (TES), which performed well in the first experiment. SFT was faster than TES across all document types and lengths, particularly when tasks involved revisitation. In addition, SFT was strongly preferred by participants.

Comparing Automatic and Manual Zooming Methods for Acquiring Off-Screen Targets

Previous studies indicate that user performance with scrolling can be improved through Speed-Dependent Automatic Zooming (SDAZ), which automatically couples the documentes zoom-level with scroll-speed. These studies have compared traditional scrolling techniques (scrollbars and rate-based scrolling) with SDAZ, leaving a potential confound that the efficiency gains are due to zooming rather than the automatic binding of zoom-level with speed. It is therefore possible that decoupling zoom from speed, allowing users separate but concurrent control of each, could further enhance performance. This paper describes an experiment (n=35) that examines user performance, workload, and preference in tasks that involve scroll-based acquisition of off-screen targets using SDAZ and manual zooming. Three different types of document navigation are explored: text documents, 'flat' 2D maps, and a 'globe browser' that allows multi-level zooming of a globe-map of Earth and underlying city views. Results show that automatic zooming not only improves performance, but that it does so with substantially less subjective workload, and that it is strongly preferred. We also confirm limited previous work using Fitts' Law as a model for off-screen target acquisition, and show that it applies even when zooming is employed.

Tuning and Testing Scrolling Interfaces that Automatically Zoom

Speed dependent automatic zooming (SDAZ) is a promising refinement to scrolling in which documents are automatically zoomed-out as the scroll rate increases. By automatically zooming, the visual flow rate is reduced enabling rapid scrolling without motion blur. In order to aid SDAZ calibration we theoretically and empirically scrutinise human factors of the speed/zoom relationship. We then compare user performance with four alternative text-document scrolling systems, two of which employ automatic zooming. One of these systems, which we term 'DDAZ', is based on van Wijk and Nuij's recent and important theory that calculates optimal pan/zoom paths between known locations in 2D space. van Wijk and Nuij suggested that their theory could be applied to scrolling, but did not implement or test their formulaic suggestions. Participants in our evaluation (n=27) completed scrolling tasks most rapidly when using SDAZ, followed by DDAZ, normal scrollbars, and traditional rate-based scrolling. Workload assessments and preferences strongly favoured SDAZ. We finish by examining issues for consideration in commercial deployments.

Speed-Dependent Automatic Zooming

Speed-dependent automatic zooming couples the user's rate of motion through an information space with the zoom level---the faster the user moves the 'higher' they fly above the work surface. Igarashi and Hinckley (2000) proposed using the technique to improve scrolling through large documents (see Takeo Igarashi's speed-dependent zooming page). Their informal preliminary evaluation showed mixed results with participants completing scrolling tasks in roughly the same time, or more slowly, than when using traditional methods. In this paper, we describe the implementation and formal evaluation of two rapidly interactive speed-dependent automatic zooming interfaces. The ecologically oriented evaluation shows that scrolling tasks are solved significantly faster with automatic zooming in both text document and map browsing tasks. Subjective preferences and workload measures also strongly favour the automatic zooming systems. Implications for the future of scrolling interfaces are substantial, and directions for further work are presented.

Click on the images below to watch brief small videos of the technique (videos both less than 1 Megabyte).

Hidden Messages: Evaluating the Efficiency of Code Elision in Program Navigation

Text elision is a user interface technique that aims to improve the efficiency of navigating through information by allowing regions of text to be `folded' into and out of the display. Several researchers have argued that elision interfaces are particularly suited to source code editing because they allow programmers to focus on relevant code regions while suppressing the display of irrelevant information. There is, however, a lack of empirical evidence of the technique's effectiveness. This paper presents an empirical evaluation of source code elision using a Java program editor. The evaluation compared a `flat text' version of this interface with two versions that diminished elided text to levels that were `just legible' and `illegible' (extremely small text). Subject performance was recorded in four tasks involving navigation through programs. Results show that programmers were able to complete their tasks more rapidly when using the elision interfaces, particularly in larger program files. Although the subjects indicated a strong preference for the just legible elision interface, their performance was best with the illegible elision system.

Tailoring Program Visualisations With Literate Programming and Fisheye Views.

Literate Tools for Introductory Programming.

Literate programming is a powerful technique that helps expert programmers integrate code and documentation in a manner that assists human comprehension. To date, tools for literate programming have had moderately complex syntactic requirements. Consequently, the programmers who stand to gain most from the clarity of the literate technique---novice and learning programmers---have been unable to do so. This paper describes the potential benefits of literate programming environments within introductory programming courses. Design criteria for such environments are presented, and prototype systems demonstrating the criteria are described. Evidence of student enthusiasm for graphical user interfaces for literate programming is discussed.

An Investigation of Groupware Support for Collaborative Awareness Through Distortion-Oriented Views

This paper reviews models and theoretical frameworks of collaborative awareness in the use of real-time groupware systems. The review is used to motivate and guide an investigation of distortion-oriented mechanisms for supporting collaborators' fluid and dynamic awareness requirements. We describe our development and evaluation of Dome, a distortion-oriented multi-user editor. Although we designed Dome to provide a realistic and useful platform for the investigation of awareness concepts, our evaluation revealed major flaws in its support for distortion-oriented awareness. We analyse the cause of these errors, some of which have been undetected in prior work, and provide precise formulations that will overcome them.

Using Distortion-Oriented Displays to Support Workspace Awareness