Impact of Visible Facial Differences on Emotional Perception

Abstract

This study looked into the impact of facial deformity on emotional perception. Our prediction that deformity could impede the recognition of emotions was based on our development of holistic face processing theories, emphasis on differences, and biased presumptions. On pictures of women with and without cleft lips, participants rated the perceived intensity of the following emotions: fear, disgust, anger, happiness, sadness, and neutral. Our research, based on data from sixty participants, shows some support for our hypotheses and may indicate a decrease in the accuracy of identifying happiness and sadness on deformed faces. Additional research will confirm these results using a more significant sample, which will also shed light on the underlying cognitive processes.

Literature Review

Studies by Barrett et al. (2019) and Boutsen et al. (2022) illuminate the complexities of deciphering emotions and altered facial processing mechanisms among individuals with facial disfigurement (FSD). These findings emphasize the challenges in accurately inferring emotions solely from facial movements, raising questions about the precision of emotional interpretation in individuals with visibly different faces. Furthermore, Jamrozik et al. (2019) emphasize the necessity of transcending superficial evaluations in societal judgments of FSD, advocating for a deeper understanding of the complexities involved in perceiving emotions and forming judgments based on facial appearance.

Despite these significant contributions, a considerable gap persists in comprehending emotional inference and societal judgments concerning individuals with visible facial differences. This research seeks to bridge this gap by examining the intricate interplay between emotional perception and societal judgments of individuals with visible facial differences. Notably, this review introduces Stone and Fisher’s (2020) brief intervention as a prospective tool for mitigating stigma associated with facial disfigurement, opening new avenues for future research and intervention strategies to alleviate societal biases and improve perceptions of individuals with visible facial differences.

Rationale For This Study

Our ability to interpret emotions from facial expressions is fundamental to social interactions. However, this innate skill encounters complexities when individuals present facial disfigurements. This study endeavors to navigate the intricate relationship between disfigurement and emotion perception, exploring a multifaceted landscape shaped by holistic processing, attention biases, and the pervasive influence of preconceived societal notions.

The human brain is inherently wired for holistic face processing, seamlessly integrating facial features to discern emotional cues (Barrett et al., 2019). Nevertheless, the presence of a facial disfigurement disrupts this intricate process. Attentional biases toward the anomaly may overshadow subtle emotional shifts, obstructing accurate recognition of emotions like happiness or sadness (Boutsen et al., 2022). Additionally, our propensity to focus on deviations from the norm can exacerbate this effect, causing a fixation on the disfigurement, thereby compromising the nuanced interpretation of emotional expressions (Jamrozik et al., 2019).

Beyond these cognitive aspects, societal biases significantly influence our perceptions of individuals with facial differences. Preconceived negative stereotypes can subconsciously affect how we interpret their emotional expressions, leading to misinterpretations and misunderstandings (Lyford-Pike & Nellis, 2021). This complex interplay of disrupted cognitive processing, attention biases, and internalized societal biases presents a formidable challenge in accurately decoding emotions on disfigured faces.

Understanding these intertwined factors extends beyond academic interest. It holds profound implications for comprehending how individuals with facial differences navigate social interactions (Posnick et al., 2019). This research seeks to develop interventions to reduce stigma and promote inclusivity by exploring the barriers and facilitators to accurate emotion recognition. Moreover, this study contributes to a broader comprehension of how facial features shape our perceptions, potentially informing various domains such as psychology, communication, and even the development of artificial intelligence capable of interpreting emotions across diverse faces.

Through this study, we aim to illuminate the intricacies of emotion perception in the context of disfigurement and to bridge the gap between individuals with uncommon facial features and broader society. By comprehensively examining the cognitive, social, and emotional dimensions, we aim to cultivate empathy, dismantle barriers, and foster a more inclusive and embracing societal landscape (Barrett et al., 2019).

Variables and Hypothesis

This study employed a mixed within- and between-subjects design to explore the impact of facial disfigurement on the perception of sadness. The 60 participants viewed six faces, each displaying sadness within each type (typical or visibly different), with three for each face type. While other emotions like fear and anger were tested as well, this study focuses on the perception of one emotion, sadness, and how it is impacted by factors such as closeness and face type. The variables were as follows;

Independent Variables (IVs)

• Facial Emotion: Sadness presented six times to each participant (within-subjects).
• Face Type: Participants were randomly assigned to view either typical or visibly different faces within each set of six (between-subjects).
• Closeness to Someone with a Visibly Different Face: Three levels (see regularly, see occasionally, none) assessed through a pre-experiment questionnaire (between-subjects).

Dependent Variable (DV)

• Perceived Sadness Intensity: Measured on a 7-point scale (1 = not at all sad, 7 = very much sad) for each sadness expression on both typical and visibly different faces.

Hypotheses

• Primary Hypothesis: Individuals viewing faces with visible differences will rate sadness expressions as significantly less intense compared to those viewing typical faces. This predicts a main effect of Face Type on perceived sadness intensity, suggesting that disfigurement might hinder accurate emotion recognition for sadness.
• Other-Hypotheses:
  • Hypothesis 1: Participants will rate sadness on visibly different faces as less accurately perceived compared to typical faces. This predicts a lower average score for the “visibly different” group on the sadness intensity scale, indicating increased difficulty in judging the emotion when a disfigurement is present.
  • Exploratory Hypothesis: The presence of a visible difference might subtly influence attention patterns towards emotional cues in the face, potentially impacting the perception of sadness. This highlights the potential for future analyses exploring eye-tracking data or other measures of attention allocation.

Method

Participants

Sixty participants voluntarily engaged in an online survey accessible through Qualtrics. Participants were recruited through university channels, including invitations during lectures, practical classes, and dissemination by enrolled students. The recruitment aimed at individuals over 18 years old who could spare approximately 10 minutes for the survey. Demographic details regarding age, gender, ethnicity, educational level, and familiarity with individuals with visible facial differences were collected.

Design

This study adopted a within-participants experimental design that concentrated on evaluating a single emotional expression, namely “sadness,” across two categories of faces: “typical” and “visibly different.” The experimental design aimed to examine and compare the perceived intensity of “sadness” depicted on facial images characterized by typical features versus those exhibiting visible differences. By focusing on a singular emotional expression, the design enabled a focused investigation into how facial distinctions might influence the perception of specific emotions, particularly the recognition of “sadness” in diverse facial presentations (Lyford-Pike & Nellis, 2021).

Materials

Facial images utilized in this study were sourced from an online database, encompassing diverse facial expressions across different age categories- young, middle-aged, and older individuals. To maintain consistency and streamline the study’s design, all images depicted female subjects. Each facial image underwent meticulous editing to simulate a cleft lip, resulting in two distinct versions of each photograph. This modification aimed to create one version displaying a typical facial appearance and another specifically altered to represent a visible difference akin to a cleft lip. The selected images covered a spectrum of age groups and ensured uniform representation across the emotions studied in the experiment. This approach facilitated the investigation of emotion perception, particularly the expression of sadness, across various age categories and facial conditions, thereby contributing to a comprehensive understanding of emotional recognition concerning visible facial differences.

Procedure

Participants engaged in an online survey where they were presented with a sequence of facial images, one at a time, depicting either a “typical” face or a “visibly different” face expressing the emotion of “sadness.” Each image was followed by a prompt asking participants to rate the intensity of “sadness” displayed on the face using a 7-point Likert scale. The Likert scale ranged from 1 (“not at all sad”) to 7 (“extremely sad”). The study focused solely on assessing the perceived intensity of “sadness” across the two types of faces presented, ensuring a targeted exploration of emotional perception regarding specific facial expressions and differences.

Data Analysis

Using IBM SPSS, a two-way Analysis of Variance (ANOVA) was performed to explore the influence of “Face Type” (categorized as “Typical” and “Visibly Different”) on participants’ perceived levels of sadness in facial expressions. The dependent variable measured was the perceived strength of sadness, rated on a Likert scale ranging from 1 (not at all) to 7 (very much). Assumptions necessary for conducting the Repeated Measures ANOVA, including assessments for normality and sphericity, were diligently checked to ensure the validity of the analysis. Once the assumptions were met, the Repeated Measures ANOVA was executed to investigate potential main effects associated with face type on perceived levels of sadness.

Discussion

Fulfillment of Hypotheses

The data mainly confirmed the study’s hypotheses. In particular, the primary hypothesis that sadness expressions on faces with obvious differences are harder to detect was validated. While the first hypothesis, that there would be a lower accuracy in identifying sadness on distinctly distinct faces, was validated, the results of the second hypothesis, stating that attention patterns would have an impact on the experience of sadness, were less clear. In spite of this, the study offered valuable insights on the identification of emotions on a variety of faces.

Comparison of Literature

The results of this study both support and contradict those of earlier studies. In line with previous research (Barrett et al., 2019; Boutsen et al., 2022; Posnick et al., 2020), our main hypothesis was supported by the strong impact that observable variations had on emotion recognition. Our exploratory hypothesis, in contrast to Jamrozik et al. (2019), indicated complex patterns in attention allocation toward emotional signals, suggesting the necessity for more in-depth research into this complex relationship.

Results

A two-way repeated measures ANOVA was conducted (Table 1), examining the effects of face type (within-subjects: typical vs. visibly different) and closeness to someone with a facial difference (between-subjects: see regularly, occasionally, never) on sadness ratings (SAD and SAD.vd) (Tables 2-3). Mean SAD ratings were higher for typical faces (M = 5.35) than visibly different faces (M = 4.87) (Table 3).

Table 1

Table 2

Table 3

Mean SAD ratings were similar across closeness levels for typical faces but higher for those who never see someone with a facial difference for visibly different faces (Figure 1).

Figure 1

Comparison of Means Between and Within Subjects

There was a significant main effect of face type on sadness ratings (Table 4, Pillai’s Trace F(1, 57) = 23.069, p < .001, η² = .288). There was a significant interaction between face type and closeness (Table 4, Pillai’s Trace F(2, 57) = 7.640, p = .001, η² = .211).

Table 4

Table 5

The results demonstrate a significant linear effect for “FaceType” (F(1, 57) = 23.069, p < .001, Partial Eta Squared = .288), indicating systematic differences in the measured variable across the levels of facial presentation. Moreover, the linear effect of the interaction between “FaceType” and “howclose” is also significant (F(2, 57) = 7.640, p = .001, Partial Eta Squared = .211), suggesting a combined linear influence of these factors on the perceived sadness.

Table 6

Table 7 presents the outcomes of the between-subjects effects analysis for the variable denoted as “MEASURE_1.” The intercept exhibits a highly significant effect (F(1, 57) = 4580.396, p < .001, Partial Eta Squared = .988), indicating substantial differences in the average outcome across the groups. Furthermore, the variable “howclose” demonstrates a significant effect (F(2, 57) = 4.031, p = .023, Partial Eta Squared = .124), suggesting differences in the perceived sadness based on varying levels of familiarity.

Table 7

The outcomes of the analyses supported the primary hypothesis, indicating differences in perceived sadness intensity between “Typical” and “Visibly Different” faces. Additionally, the interaction effect suggested that familiarity with facial differences influences how sadness is perceived across different face types, offering valuable insights into emotional recognition concerning facial distinctions.

Effect Size and Assumptions

Table 8 below shows the Shapiro-Wilk tests, exceptionally reliable for smaller sample sizes, which indicate that SAD.vd likely adheres to a normal distribution across all “how close” levels. Regarding SAD, it appears to conform to a normal distribution within the “see regularly” and “never” groups, although a marginal deviation from normality arises within the “occasionally” group. Additionally, the Kolmogorov-Smirnov test with Lilliefors correction suggests a potential non-normality for SAD.vd in the “occasionally” group (p = 0.010). Nevertheless, the Shapiro-Wilk results are prioritized for assessing normality in this context due to its decreased sensitivity. Mauchly’s Test of Sphericity was conducted for the within-subjects effect measured by “Perception of Sadness,” mainly focusing on the variable “FaceType.” The test resulted in a Mauchly’s W of 1.000 (approximate chi-square = .000, df = 0), indicating that the assumption of sphericity was met, implying that the error covariance matrix for the transformed dependent variables is proportional to an identity matrix (Table 9).

Table 8

Table 9

Implications and Applications

This work clarifies the important consequences of visible facial variations in social interactions, especially with relation to the identification of emotions. Our results demonstrate that facial disfigurements can make it difficult to accurately perceive certain emotions, such as sadness (Posnick et al., 2020). This indicates that people with unusual features may be more likely to misunderstand and misinterpret others in ordinary situations. According to Boutsen et al. (2022), the fundamental mechanism is based on attentional biases against disfigurements that obscure minor emotional cues. In order to effectively connect with people who have face variations, it is imperative that one exercise greater sensitivity and awareness.

Furthermore, the way that familiarity and emotion perception interact offers intriguing new perspectives on how flexible our emotional interpretations might be. Our perception of emotions on unique faces is greatly influenced by our exposure to varied facial traits and our personal experiences (Jamrozik et al., 2019). This provides opportunities for future therapies that attempt to improve the ability to recognize emotions and lessen societal prejudices. Initiatives to promote social contact and empathy training have the potential to create an environment that is more accepting and understanding of everyone. Our research reveals connections between cultural norms, emotion perception, and facial traits, which have practical implications for psychology and communication. This motivates culturally sensitive and accurate inclusive interventions. It also steers Artificial Intelligence (AI)’s bias-resistant emotion identification algorithms, improving justice in practical uses like healthcare and education and guaranteeing interactions that are considerate of cultural differences (Lyford-Pike & Nellis, 2021).

Limitations and Further Research

The study’s limitations include a small sample size, potentially constraining result generalization. Future studies should aim for larger, diverse samples to validate and broaden these findings. Additionally, the sole focus on sadness perception limits a comprehensive grasp of emotional recognition. Examining other emotions would enhance understanding in disfigurement-related emotional processing. Moreover, using static face images might limit capturing emotional dynamics; exploring dynamic stimuli like video recordings could offer deeper insights.

Conclusion

Our study concludes by highlighting the significant influence of face variations on the perception of emotions, particularly sadness. It highlights the difficulties experienced by people with obvious facial differences and the urgent need for initiatives that promote empathy and understanding in social situations. Even with its small sample size and exclusive attention to the perception of sadness, this work establishes the foundation for understanding the complex relationship between facial appearance and emotional recognition. Our results, which we arrived at using IBM SPSS’s two-way ANOVA function, offer a vital starting point for more investigation into this intricate link and call for the creation of policies that will promote inclusivity and social understanding.

References

Barrett, L. F., Adolphs, R., Marsella, S., Martinez, A. M., & Pollak, S. D. (2019). Emotional expressions reconsidered: Challenges to inferring emotion from human facial movements. Psychological science in the public interest, 20(1), 1-68.

Boutsen, L., Pearson, N. A., & Jüttner, M. (2022). Do facially disfiguring features influence attention and perception of faces? Evidence from an antisaccade task. Quarterly Journal of Experimental Psychology, 75(5), 830–840.

Jamrozik, A., Oraa Ali, M., Sarwer, D. B., & Chatterjee, A. (2019). More than skin deep: Judgments of individuals with facial disfigurement. Psychology of Aesthetics, Creativity, and the Arts, 13(1), 117–129. https://doi.org/10.1037/aca0000147

Lyford-Pike, S., & Nellis, J. C. (2021). Perceptions of patients with facial paralysis: predicting social implications and setting goals. Facial Plastic Surgery Clinics, 29(3), 369–374.

Posnick, J. C., Ogunsanya, O., Singh, N., & Kinard, B. E. (2020). Short face dentofacial deformities: changes in social perceptions, facial esthetics, and occlusion after bimaxillary and chin orthognathic correction. Journal of Craniofacial Surgery, 31(3), 632–636.

Posnick, J. C., Susarla, S. M., & Kinard, B. E. (2019). Reconstruction of residual cleft nasal deformities in adolescents: effects on social perceptions. Journal of Cranio-Maxillofacial Surgery, 47(9), 1414-1419.