[Features of emotional voices: focus in differences between expression and recognition]

Area of Science:

• Psychology
• Linguistics
• Speech Science

Background:

• Accurate emotion recognition from speech is crucial for human interaction.
• Previous research has explored vocal emotion expression and perception, but the interplay between acoustic cues and perceived emotional categories requires further investigation.

Purpose of the Study:

• To investigate the relationship between emotions expressed vocally and how they are perceived by listeners.
• To identify specific emotions that are frequently misjudged and explore the reasons behind these misinterpretations.
• To examine the role of acoustic cues and emotional category similarities in vocal emotion judgment.

Main Methods:

• Study 1: 21 participants rated emotions expressed by an actress, with an average accuracy of 67.6%.
• Study 2: 25 participants judged emotions from 23 speakers, achieving an average accuracy of 48.6%.
• Speech analysis was used to examine similarities in vocal features across different emotions.

Main Results:

• Judgments of vocal emotions were not always accurate, with specific emotions being more prone to misinterpretation.
• Misjudgments often occurred between emotions with similar categories (e.g., sadness and relaxation) or similar vocal features.
• Acoustic analysis revealed shared vocal characteristics among certain emotions, contributing to perceptual confusion.

Conclusions:

• The accuracy of vocal emotion recognition is influenced by both the acoustic properties of the voice and the perceived similarity between emotional categories.
• Similarities in vocal features and emotional categorization can lead to systematic errors in emotion judgment from speech.
• Future research should consider these factors when developing models for automatic emotion recognition.