A laboratory-based procedure for measuring emotional expression from natural speech

Area of Science:

• Psychology
• Linguistics
• Computer Science

Background:

• Measuring emotional expression in speech is crucial for understanding psychological disorders.
• Existing tools face methodological challenges in simultaneously analyzing prosodic and content speech features.
• Simultaneous measurement is vital for laboratory research on emotion.

Purpose of the Study:

• To evaluate the Computerized assessment of Affect from Natural Speech (CANS) for laboratory research.
• To assess CANS's ability to measure lexical and prosodic expression across varied emotional conditions.
• To explore the interface between lexical and prosodic expressions in emotional responses.

Main Methods:

• Digitally recorded verbal responses from 69 participants (38 male, 31 female).
• Participants reacted to pleasant, unpleasant, and neutral emotional stimuli.
• Analysis of lexical (content) and prosodic (vocal tone) expression variables.

Main Results:

• Significant changes in lexical and prosodic expression variables were observed across conditions.
• The CANS procedure demonstrated effectiveness in differentiating emotional states through speech.
• Findings support the utility of CANS for laboratory-based emotional expression research.

Conclusions:

• The Computerized assessment of Affect from Natural Speech (CANS) is a viable tool for laboratory research.
• CANS allows for simultaneous measurement of lexical and prosodic emotional cues.
• Further research can utilize CANS to investigate emotional expression and its disruption in psychological disorders.