Michele Masterson

Researchers Examine How Characteristics of Automated Voice Systems Affect Users’ Experience

The personality and gender of the automated voices you hear when calling your credit card company or receiving directions from your GPS navigational system may have an unconscious effect on your perception of the technology.

Researchers Rochelle Edwards and Philip Kortum at Rice University have conducted a study focusing how the gender and tone selected for an IVR affects its user-friendliness in which participants interacted with a medical IVR that collected information about their health. Users responded to both male and female voices that spoke in different tones, such as upbeat, professional, or sympathetic, and were then asked to judge the system’s usability. Edwards and Kortum will present their paper, “He Says, She Says: Does Voice Affect Usability?” at an upcoming Human Factors and Ergonomics Society meeting,

Past studies have indicated that users are more responsive to actual human voices than to computer-generated voices, but little research has been completed on the role that voice characteristics play in user perceptions of the technology.

“We have been systematically looking at what affects user performance on IVRs for some time now,” said Kortum. “Voice is the major element in an IVR interface, as graphical elements are for a Web page, and this study was a first attempt to understand the impact voice might have on the perceived usability of such systems.”

The authors found that although IVRs with male voices tended to be perceived as more usable than those with female voices, they were not considered more trustworthy. The researchers encourage designers to take voice characteristics into consideration when developing future systems.

“Anyone who uses an IVR knows how frustrating they can be,” said Kortum. “Much of this frustration stems from poorly designed IVRs, not from the form of interface being intrinsically “bad.” This research shows that some simple modifications to the design of these systems can have an impact on the usability of voice interfaces.”

Source: The Human Factors and Ergonomics Society