Children’s ratings of vocal emotion intensity depend on the emotion spoken and speaker familiarity but not acoustic parameters

Abstract

Scherer (1986) documented the acoustic parameters associated with adults’ perceptions of discrete vocal emotions. We investigated physical and psychological factors that influence children’s ratings of discrete vocal emotions using stimuli that approximated naturally occurring speech, including familiarity with the speaker. Specifically, we presented 52 7- and 8-year-olds with one side of a brief phone conversation spoken in happy, angry, sad, and non-emotional prosodies by both the child’s mother and another child’s mother, unfamiliar to the target child. As a group, the familiar and unfamiliar mothers’ prosodies did not differ in fundamental frequency (F0), F0 standard deviation, or speech rate—acoustic parameters that are most identified with angry, happy, and sad prosodies. Children accurately recognized the emotion spoken: They rated angry stimuli as more angry than happy or sad. Regression analyses indicated that speaker familiarity predicted children’s intensity ratings even after the target acoustic parameters were taken into account, but this effect was moderated by speaker emotion, such that children rated their mothers as more intensely angry than unfamiliar mothers. The findings suggest that their mother’s angry voice holds psychological significance for children that is not explained by variations in its most salient acoustical properties.