Emotion and motion: age-related differences in recognizing virtual agent facial expressions

Abstract

Technological advances will allow virtual agents to increasingly help individuals with daily activities. As such, virtual agents will interact with users of various ages and experience levels. Facial expressions are often used to facilitate social interaction between agents and humans. However, older and younger adults do not label human or virtual agent facial expressions in the same way, with older adults commonly mislabeling certain expressions. The dynamic formation of facial expression, or motion, may provide additional facial information potentially making emotions less ambiguous. This study examined how motion affects younger and older adults in recognizing various intensities of emotion displayed by a virtual agent. Contrary to the dynamic advantage found in emotion recognition for human faces, older adults had higher emotion recognition for static virtual agent faces than dynamic ones. Motion condition did not influence younger adults' emotion recognition. Younger adults had higher emotion recognition than older adults for the emotions of anger, disgust, fear, happiness, and sadness. Low intensities of expression had lower emotion recognition than medium to high expression intensities.