Poster #114 - Speaker Familiarity Influences Children's Perceptions of Vocal Emotional Intensity, but not Acoustic Parameters

• In Event: 3-159 - Poster Session 12
  In Poster Session: PS 12 Section - Language, Communication

Integrative Statement

Children’s accurate interpretations of vocal emotion cues are included among the essential components of socioemotional competence (Halberstadt, Denham, & Dunsmore, 2001; Saarni, 1991). School-aged children accurately label emotions based on affective prosody (Berman, Chambers, & Graham, 2010), but understanding others’ emotions is about more than just emotion labeling. Speech interactions in children’s daily lives are more dynamic, involving the need to process the personal relevance of the speaker, the emotion expressed, and how intensely that emotion is expressed. Our poster presents evidence of 7- and 8-year-olds’ intensity ratings of speakers’ emotions, contrasting a familiar speaker (mother) with an unfamiliar speaker (another child’s mother who is not familiar).
We investigated the acoustical and perceptual factors that influence children’s ratings of vocal emotional intensity using stimuli designed to mimic naturally occurring speech. Specifically, 52 7- and 8-year-olds listened to speech samples spoken by both their mothers and other, unfamiliar mothers. The speech stimuli simulated one side of a phone conversation and each script, written to be emotionally neutral, was spoken in four prosodies: angry, happy, sad, and non-emotional prosodies. At the group level, the acoustical qualities of familiar and unfamiliar mothers’ prosodies did not differ in terms of fundamental frequency (F0), F0 standard deviation, speech rate, or energy, all ps > .606. These acoustic parameters distinguish angry, happy, and sad prosodies (Scherer, 1996). In contrast, children rated their own mothers as more intensely emotional than unfamiliar mothers, F(1, 50) = 12.02, p = .001, ƞp2 = .16, 95% CI [0.07, 0.25], an effect moderated by the specific emotion used. Children rated their mothers’ speech as more intensely angry, F(1, 50) = 14.25, p < .001, ƞp2 = .22, 95% CI [0.13, 0.42], and happy, F(1, 50) = 6.41, p = .015, ƞp2 = .11, 95% CI [0.05, 0.40], than unfamiliar mothers’ speech. Hierarchical regression analyses confirmed that the speaker familiarity for angry speech accounted for a significant amount of variance over and above the acoustic parameters, B = .14, t(96) = 2.64, p = .010, 95% CI [0.03, 0.24]. The effect of speaker for happy speech was not significant in the hierarchical regression analyses, B = .11, t(96) = 1.61, p = .111, 95% CI [-0.03, 0.24]. Further, the acoustic properties as a set were not significantly associated with children’s ratings of emotional intensity for any of the emotion categories, all ps > .266. The acoustic properties did not predict children’s intensity ratings as we expected given the adult literature (Scherer, 1996), but this suggests that there may be other factors contributing to children’s perceptions of emotional intensity. However, these findings suggest that their mother’s voice holds psychological significance for children, particularly for angry speech as has been demonstrated in prior literature (Shackman & Pollak, 2005). The present study expands upon this literature, however, by highlighting these differences are entirely perceptual, and that the effect is not attributable to the acoustical properties that determine affective prosody.

Authors

• Tawni Blaze Stoop, Pennsylvania State University
  Presenting Author

• Peter Moriarty, Pennsylvania State University
  Non-Presenting Author

• Michelle C Vigeant, Pennsylvania State University
  Non-Presenting Author

• Rick O Gilmore, Pennsylvania State University
  Non-Presenting Author

• Pan Liu, Pennsylvania State University
  Non-Presenting Author

• Pamela M. Cole, The Pennsylvania State University
  Non-Presenting Author