| Sumario: | Purpose A complex interface with several controls and multiple choices, such as an interface of a washing machine, would be hard to recognize and manipulate without visual information and feedback. People with visual disabilities are always exposed to this type of difficulty when using such interfaces. This study examines the informational structure of interfaces of various electronic appliances in terms of arrangement width and depth. This is expressed by multiplying the number of controls by the number of choices. This study then reveals how the structure affects the usability of the interface. We examined this usability question: "Will the canonicity (information structure in the canonical form) affect learning and performance of a person attempting to locate a choice from a series of multiple control options particularly when there is no visual information but only with auditory feedback?" Method The choices can be grouped under each control in two ways: canonically balanced and random. The canonically balanced arrangements of 3x5 (three controls with five choices each) and 5x3 (five controls with three choices each) were used in our study. The random and non-balanced arrangement does not have a canonical arrangement: three controls with three choices for the first control, five choices for the second control, and four choices for the last. The usability was assessed using the learning time (actually time to memorize the structure), and time to locate a given choice for 30 young blindfolded subjects with a prototype control panel that has three to five push buttons. Results & Discussion The learning time and locating time increased as the number of controls and the choices increased (Figure 4). Performance improved when the canonicity was maintained and when the amount of the information to transmit was equal between the balanced and random arrangements (Figure 5). Within the canonically balanced structure, the interface with fewer controls and more choices in a control are better than the opposite structure in learning and locating times. In general, user performance was better when canonicity of the information structure is maintained, and when information is densely chunked within a control. The study can provide guidelines for user interface designs of various household appliances with controls for the blind users and older users. For some users who have the short-term memory problems, such as the users with cognitive impairment or older users, auditory feedback seemed to not convey as much information as visual information does for learning where the option to select is located for the desired function. Even without auditory feedback, the interface needs to be designed with canonicity maintained in information structure in order to provide accessibility to the users with visual impairment or memory problems.
|
|---|