NSF-Funded Student Design Projects:
Fitts' Law Game System for Teaching Accessible Design

Designers: Vamsi Ramakrishna Penmecha
Client Coordinators: Dr. Robert Erlandson, Enabling Technologies Lab
Supervisor: Dr. Robert Erlandson
Department of Electrical and Computer Engineering
Wayne State University
Detroit, MI 48202

INTRODUCTION
There are many dimensions to accessible design and  Fitts’ Law is one of many scientific laws that influences accessible design strategies. Current federal legislation mandates the use of accessible design in a variety of services and products, yet a large number of university faculty and hence students are generally not aware of these legal mandates. The Fitts’ Law Game System is designed to raise student and faculty awareness of accessible design principles in general and Fitts’ Law in particular.

SUMMARY OF IMPACT
In addition to raising awareness of accessible design issues, the system serves two other purposes in illustrating Fitts' Law as well. One, teachers and therapists use a variety of augmentative communication devices for their special education students and clients.  These devices typically use a series of overlays that the users touch to produce a message or computer action if the device is used as an alternative keyboard. The overlay design is a specific example of the application of Fitts’ Law, and hence overlay designers need also to be aware of the implications of Fitts’ Law on the design of their overlay products. Second, the mathematical expression of Fitts’ Law was developed using the general population as experimental subjects. The Fitts’ Law Game System will allow data collection for a variety of users. This will enable data to be gathered for individuals with disabilities in order to test the validity of Fitts’ Law for diverse user populations.

TECHNICAL DESCRIPTION
Fitts' Law basically states that the bigger the target, the easier it is to hit.  The law can be described by an equation that specifies the relationship between movement time (MT), movement distance (D), target size (A) and various characteristics of the person, method and task modeled by two parameters K1 and K2.  It has the following basic form:

MT=K1 + K2log(2D/A)
 

There have been many refinements and applications of this basic equation.  Part of the appeal of the equation is that it is intuitive for those who understand the mathematics.  The further you move (D), the longer it takes.  The bigger the target, the shorter the target acquisition time (MT).  The non-linear, logarithmic equation (model of human performance) is typical of  models describing human performance.

Examples of Fitts’ Law can be found in push-button TV and VCR remote controllers. The size and placement of the push buttons make operation more or less difficult. Other examples include the push-button controls on car radios and light switches.  Processes such as placing groceries into a shopping bag, inserting an ATM card into the receiving slot, and pulling a car into the garage all exhibit principles of Fitts' Law.  The assembling of components on an automobile has been explored extensively with the slogan “Tight Targets Take Time” applied to the very precise timing of high volume assembly operations and the design opportunities that exist to create bigger targets.

A universal example of Fitts’ Law is a computer’s cursor control via a mouse or other input device. Because this is such a fundamental task the Fitts’ Law Game System is built around cursor control on a computer monitor. The player can select one of six levels of difficulty (See Figure 1 below).  The lower the difficulty, the larger the target area.  The higher the level of difficulty, the smaller the target’s area. The system always brings the cursor to a home position in the lower right hand corner of the screen. A given target is then randomly presented on the screen.  The player must move the cursor into the target area and click the mouse, or perform an equivalent operation that signifies a mouse click if using an alternative input device.

There are six different target areas and ten different distances. For a given target, one of the distances (a direct line from the home position to the target) is randomly generated. A given distance falls along an arc with the home position as the center of the arc.  The system generates a random position along the arc.  This scheme prohibits placement patterns yet provides data for fixed target areas and specified distances.

A player can specify the number of targets that will be presented and can play the game six times per setting, one game for each target size or level of difficulty. The following data is automatically collected:  target area, distance moved, and movement time (time from the appearance of the target until the user clicks the mouse within the target area). When the player is done, the system will provide an option to see a plot of the results.  The results can be displayed in one of the following ways:  movement time versus target area (with distance as a parameter) or movement time versus distance (with target area as a parameter).

The system is designed as a game, but can be configured to formally gather performance data.  In the experimental (data collection) mode the experimenter identifies the subject via a code. The experimenter selects the difficulty level and number of targets to be presented. Data is automatically collected and saved into a subject data file.

No costs were incurred in developing this project as it was designed using currently available resources and equipment.