Facing threat: Infants' and adults' visual scanning of faces with neutral, happy, sad, angry, and fearful emotional expressions

Abstract

Human faces are among the most important visual stimuli that we encounter at all ages. This importance partly stems from the face as a conveyer of information on the emotional state of other individuals. Previous research has demonstrated specific scanning patterns in response to threat-related compared to non-threat-related emotional expressions. This study investigated how visual scanning patterns toward faces which display different emotional expressions develop during infancy. The visual scanning patterns of 4-month-old and 7-month-old infants and adults when looking at threat-related (i.e., angry and fearful) versus non-threat-related (i.e., happy, sad, and neutral) emotional faces were examined. We found that infants as well as adults displayed an avoidant looking pattern in response to threat-related emotional expressions with reduced dwell times and relatively less fixations to the inner features of the face. In addition, adults showed a pattern of eye contact avoidance when looking at threat-related emotional expressions that was not yet present in infants. Thus, whereas a general avoidant reaction to threat-related facial expressions appears to be present from very early in life, the avoidance of eye contact might be a learned response toward others' anger and fear that emerges later during development.

Keywords:

• Visual scanning
• Infants
• Facial expression
• Emotional expression
• Threat

Acknowledgements

This research was supported by grants to CvH from the Tercentennial Fund of the Bank of Sweden (J2004–0511) and European Union (EU15636: TACT). TdW was funded by the Niels Stensen foundation.

We would like to thank all families who took part in the study and Dorota Stasiewicz and Angela Khadar for their assistance.

Notes

¹To further examine these effects and to substantiate the interpretation that the effect of emotional expression on the relative number of fixations and the dwell times to the inner features of the face was indeed already present in the youngest age group an additional MANOVA was calculated for the 4-month-old group. The analyses yielded a significant effect of emotion on the relative dwell time to the feature areas, F(4, 120)=3.18, p<.05, and a marginally significant effect for the number of fixations, F(4, 120)=2.06, p<.10. The pattern in the data follows the pattern as described for the entire sample, with a higher percentage of dwell time on the feature areas of non-threat-related (neutral: M=71.0%, SE=3.4%; sad: M=69.2%, SE=3.2%; happy: M=68.7%, SE=3.5%) compared to threat-related emotional expressions (angry: M=65.5%, SE=3.6%; fearful: M=62.6%, SE=3.8%) and also a higher percentage of fixations to feature areas of non-threat-related (neutral: M=68.8%, SE=3.3%; sad: M=69.2%, SE=3.2%; happy: M=67.1%, SE=3.3%) compared to threat-related expressions (angry: M=62.7%, SE=3.4%; fearful: M=63.4%, SE=3.4%; cf. Figure 2). Since post hoc t-tests did not provide sufficient statistical power to further elaborate the pattern of this effect, we instead performed an effect size analysis. For each of the five levels of the factor emotional expression we calculated the effect size parameter d _z using the G*Power 3 software (Faul, Erdfelder, Lang, & Buchner, 2007). Concerning the percentage dwell times medium-sized effects were found for the difference between neutral and fearful expressions, d _z=0.51, and sad and fearful expressions, d _z=0.48. For the differences between neutral and angry and sad and angry expressions, the effect sizes were smaller, d _z=0.32 and d _z=0.38, respectively. Concerning the relative number of fixations, an effect size of d _z=0.36 was found for the comparison of neutral versus angry expressions and of d _z=0.42 for the comparison of sad versus angry expressions. The comparison of neutral versus fearful and sad versus fearful expressions revealed effect sizes of d _z=0.30 and d _z=0.31, respectively.