Assessing collective affect recognition via the Emotional Aperture Measure

Abstract

Curiosity about collective affect is undergoing a revival in many fields. This literature, tracing back to Le Bon's seminal work on crowd psychology, has established the veracity of collective affect and demonstrated its influence on a wide range of group dynamics. More recently, an interest in the perception of collective affect has emerged, revealing a need for a methodological approach for assessing collective emotion recognition to complement measures of individual emotion recognition. This article addresses this need by introducing the Emotional Aperture Measure (EAM). Three studies provide evidence that collective affect recognition requires a processing style distinct from individual emotion recognition and establishes the validity and reliability of the EAM. A sample of working managers further shows how the EAM provides unique insights into how individuals interact with collectives. We discuss how the EAM can advance several lines of research on collective affect.

Keywords:

• Group emotion
• Emotion recognition
• Affective tone
• Global processing
• Emotional intelligence
• Social intelligence

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ EAM stimuli and programming scripts are available for non-commercial research purposes at www.tinyurl.com/emotional-aperture.

² Another advantage of the two-frame video clip approach includes minimising artefactual errors in collective affect recognition that could occur when observers are not first provided with a baseline expression (e.g., neutral expression or emotional expression different from that displayed in the second frame). For example, consider a professional cricket coach who works outdoors and has developed deep vertical lines on his forehead as a consequence of squinting in the glaring sun for hours each day. Without considering a facial baseline for this particular individual, observers may mistake him for feeling negative even when he is feeling neutral or mildly pleasant (Ekman & Friesen, 1976). The first frame of the movie clips provides this critical baseline expression.

³ The authors affirm that sample size, any data exclusions, manipulations and measures are each discussed as relevant for the studies in this research.

⁴ To provide a more direct assessment of the EAMs test-retest reliability, an independent sample of 72 participants (25 men, 47 women, M_age = 20.91, SD = 3.33) completed the EAM twice with two days between the first and second assessment. The test-retest EAM scores were significantly positively correlated, r (72) = 0.62, p < 0.01. A repeated measures ANOVA further showed a significant increase in EAM scores between the first assessment (M = .84, SD = 0.11) and the second assessment (M = .89, SD = 0.11), F(1, 71) = 18.52, p < 0.01, partial η² = .21.

⁵ In a separate study using a similar experimental design as Study 1, we examined whether individual emotion recognition would vary as a function of the global/local prime by asking participants (n = 99) to complete the DANVA before and after completing either the global or local processing prime. A mixed design ANOVA with DANVA-Time 1 and DANVA-Time 2 as the repeated measure and global/local prime as the between-subjects condition showed a prime × time interaction, F(1, 97) = 3.98, p = 0.05. There was no evidence for a significant change in DANVA score from Time 1 to Time 2 as a result of the local prime, t(47) = 0.47, p = 0.64. However, there was a significant decrease in DANVA score from Time 1 to Time 2 as a result of the global prime, t(50) = 2.30, p = 0.03, Cohen's d = 0.25. These results further highlight the implications of a global processing style for collective (helpful) vs. individual (harmful) emotion recognition (for an insightful review of how different processing styles shape affective recognition vs. facial identification, see Calder, Young, Keane, & Dean, 2000, and Ellison & Massaro, 1997).