Differences in adult and adolescent listeners’ ratings of valence and arousal in emotional prosody

ABSTRACT

Judgments of emotional stimuli’s valence and arousal can differ based on the perceiver’s age. With most of the existing literature on age-related changes in such ratings based on perceptions of visually-presented pictures or words, less is known about how youth and adults perceive and rate the affective information contained in auditory emotional stimuli. The current study examined age-related differences in adolescent (n = 31; 45% female; aged 12–17, M = 14.35, SD = 1.68) and adult listeners’ (n = 30; 53% female; aged 21–30, M = 26.20 years, SD = 2.98) ratings of the valence and arousal of spoken words conveying happiness, anger, and a neutral expression. We also fitted closed curves to the average ratings for each emotional expression to determine their relative position on the valence-arousal plane of an affective circumplex. Compared to adults, adolescents’ ratings of emotional prosody were generally higher in valence, but more constricted in range for both valence and arousal. This pattern of ratings is suggestive of lesser differentiation amongst emotional categories’ holistic properties, which may have implications for the successful recognition and appropriate response to vocal emotional cues in adolescents’ social environments.

KEYWORDS:

• Emotion
• prosody
• arousal
• valence
• adolescence
• development

Acknowledgements

We are grateful to Brad Moffat for his invaluable help and expertise.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 The interaction effect was significant if age was instead entered into the model as a continuous variable, for both valence, F(1.73, 100.46) = 3.49, p = .04, η² = .06, and arousal, F(1.41, 82.01) = 6.96, p = .005, η² = .11. Parameter estimates suggest that increased age was associated with lower ratings of valence for the younger speaker’s portrayal of anger, B = −.04, β = −.15, p = .02, η² = .09, and lower ratings of arousal for the older speaker’s portrayal of neutral, B = −.07, β = −.53, p = .03, η² = .08.

2 Results were similar if listener age was instead entered into the model as a continuous variable. Though the omnibus effect was not significant (p = .13), age predicted V_sin (p = .03) and A_sin (p = .03), but not V₀ (p = .14).