Stretch Button Scrollbar

Daniel J. Smith*, Robert A. Henning **

*Dept. of Psychology, University of Connecticut
Storrs, CT 06269
(203) 486-5557, dsmith@psych.psy.uconn.edu

**Dept. of Psychology, University of Connecticut
Storrs, CT 06269
(203) 486-5918, henning@uconnvm.uconn.edu

ABSTRACT

Previous research has examined the use of a scrollbar for item selection in very large lists. This report presents an alternative design based on an understanding of the action capabilities and perceptual feedback. Experimental results show substantial performance benefits when using the Stretch Button scrollbar compaired to other designs.

Keywords:

scrollbar, widget, selection technology, action/perception based control, dynamic system.

INTRODUCTION

Computer systems can control and present large amounts of information in complex and dynamic ways, often requiring operators to manipulate control variables containing thousands of values. This report describes a graphical user interface (GUI) design and evaluation project intended to meet several goals:

Allow a user to rapidly search a very large list of items.
Reduce required screen space to a minimum.
Use only with pointing device (mouse, pen, trackball).

Previous research attempted to achieve these goals by modifying the design of the standard scrollbar GUI element. However, empirical data analysis did not find improved performance [1]. This project attempts to achieve the above goals by taking an action/perception based control perspective on the task of item search.

Item Search Task Analysis

Item search involves browsing and/or searching a list of items for a target, and is theorized to be a dynamic, feedback controlled process incorporating the action capabilities and perceptual feedback properties of the system. According to Flach [2], the key to control over dynamic, complex activity is the integration and coordination of action and perception. In this case, the fundamental element of item search is control of position, enabled by movement within the list. While movement can be viewed as a continuum of item scanning speed ranging from gross to fine movement, it is valuable to explicitly describe the endpoints of the continuum.

Gross Movement. Very large jumps through the list.

Range of Moderate Movement. Control over position, achieved through a closed-loop feedback process.

Fine Movement. Movement by single items.

Action/Perception Based Control

The basic idea of action/perception based control has been extended to input devices, calling for a deeper analysis than simple surface structure (e.g., analysis based on Fitts' law). Optimal control requires an analysis of the perceptual structure of the task, the control (action) properties of the input device, and their interrelationship [3]. An action/perception based control perspective on design suggests a control system designed for searching a very large list of items must integrate:

The action capabilities which enable control of position using the continuum of movement.
Perceptual feedback which facilitates action regulation.

DESIGN

The design goals severely limit the possibilities for effective and flexible HCI design, however, previous research found the layout shown in Figure 3 to be an acceptable alternative to menus or keyboard input [1]. Design opportunities, in this case, center around modification of the standard scrollbar GUI element. To enable rapid item search, access to the actions which enable control over position must be provided. Thus, the following actions are to be supported: 1) Gross movement. 2) Range of moderate movement. 3) Fine movement.

To properly regulate these actions, perceptual information must not only indicate one's position within the list, but provide information that enables one to regulate one's action. The capability for closed loop action/perception based control will enable one to effectively control and coordinate action to find the target.

Stretch Button Scrollbar

This scrollbar consists of a bar which containing a thumb with three components (Figure 1). The various surfaces afford several opportunities for action to the user:

Figure 1: Stretch Button Scrollbar

Gross movement: Thumb's center button can be used to drag the thumb along the length of the bar, changing item list position relative to the thumb's position within the bar.

Range of moderate movement: Dragging upon an arrow button away from the thumb will change position within the list by one item per mouse movement (Figure 2).

Figure 2: Stretch Button Thumb

Fine movement: Clicking on either arrow button (without dragging) changes list position by a single item.

Support for Item Search: The Stretch Button scrollbar design gives the user the opportunity to perform the full range of actions required for control of position within the item list. This allows one to directly access activities appropriate for progressing towards the target position quickly and accurately. In addition, the design includes commensurate perceptual feedback for regulating action.

EXPERIMENT

To evaluate the Stretch Button scrollbar design, an experiment was conducted to compare search performance of the Stretch Button scrollbar, the Standard scrollbar design, and the Alphaslider design [1]. The experiment involved thirty three participants searching for a target word within an alphabetical list of 10,000 words.

Procedure

The within-subjects experimental design attempted to minimize order and learning effects by counter-balancing the order of design presentation, and giving the participants the opportunity to practice with each scrollbar design. The participants were given an overview of the experimental task, and instructions for using each scrollbar design. A computer program presented the experiment to each participant and recorded the performance data (Figure 3).

Figure 3: Example Computer Program Screen

For each experimental trial, the time to find the target word was recorded by the computer program. Participants took between 40 and 60 minuets to find 25 words with each design, after which they completed a twelve question subjective preference questionnaire.

Hypotheses

The Stretch Button scrollbar will enable participants to find the target word in less time than the other scrollbar designs.

RESULTS

Participants were able to find the target word, on average, in less time using the Stretch Button design than the other two scrollbar designs (Figure 4). ANOVA analysis found the differences between the three scrollbar designs to be statistically significant (F(2,69) = 8.84, p < 0.001) and planned comparison analysis found the Stretch Button scrollbar design to be significantly faster than the other two designs (p < 0.05).

Figure 4: Average Time to Find Target Word

DISCUSSION

In terms of the stated design goals the Stretch Button scrollbar is a success and an improvement on the standard scrollbar design. Analysis of the item search task using an action/perception perspective provided a basis for understanding the fundamental elements of the task Elaboration of the movement continuum uncovered three basic activities involved in item search: 1) Gross movement. 2) Range of moderate movement. 3) Fine movement.

By designing for direct access to these activities, the Stretch Button scrollbar affords a broader range of movement control than other designs. In addition, the action/perception based control perspective focused attention on the interrelationship between the action capabilities of the control and the regulating perceptual information provided by the control. The result is a design where action and perception are coordinated, improving performance over previous designs. This allows designers to provide access to many control variables simultaneously, enabling complex and dynamic system control.

Further Research

It may be possible to apply the Stretch Button scrollbar to other areas where the Standard scrollbar design is used. Only after careful consideration of the action/perception based control characteristics of the task can it be determined if this design is applicable. In addition, other common GUI control widgets may benefit from an action/perception based analysis. By concentrating on the inter-relationship between action capabilities and perceptual information, users may be able to control complex computer systems.

REFERENCES

Ahlberg, C., & Shneiderman, B. The Alphaslider: A compact and rapid selector. Proceedings of CHI'94 Human Factors in Computing Systems (Boston, April 1994), 365-371.
Flach, M. J. (1995). Beyond Error: The language of coordination and stability. In J. Rasmussen & B. Brehmer (eds.) The Evolution and Breakdown of Adaptive Systems. pp. 231. Wiley & Sons, New York.
Jacob, R. J. K., Sibert, L. E., McFarlane, D. C., & Mullen, M. P. Jr. (1994). Integrality and separability of input devices. ACM Transactions on Computer-Human Interaction, 1(1), 3-26.