Facial Affect Research Articles

Face the music: Children’s facial affect in musical imitation and improvisation tasks

The purpose of this study was to examine facial affect of young children who completed a singing task that included imitation and improvisation. Eighty-nine children (4–6 years: 45 male and 44 female participants) completed three singing conditions from a standard singing test battery (i.e., Advancing Interdisciplinary Research in Singing-Test Battery of Singing Skills [AIRS-TBSS]). These included singing a favorite song, imitating a song, and improvising a song ending. Facial affect was analyzed with Noldus FaceReader software, and subjective responses also were collected. Results revealed children exhibited a happy emotion most prominently during the improvisation and favorite song conditions compared with the imitation condition. However, a higher percentage of surprised emotions were found during the imitation condition. Frequency analysis revealed a significantly different range and final note for the improvisation condition compared with imitation. Children’s self-reported ratings of happiness were related to their displayed facial affect (i.e., happiness) scores in FaceReader ( p < .05). Qualitative data analysis revealed three emerging themes of song familiarity, object association, and song preference. Children exhibited more positive affect when singing a favorite song or improvising. Based on the type of vocal performance task, it is necessary to consider how young children respond to vocal tasks.

Orbitofrontal control of conduct problems? Evidence from healthy adolescents processing negative facial affect

Conduct problems (CP) in patients with disruptive behavior disorders have been linked to impaired prefrontal processing of negative facial affect compared to controls. However, it is unknown whether associations with prefrontal activity during affective face processing hold along the CP dimension in a healthy population sample, and how subcortical processing is affected. We measured functional brain responses during negative affective face processing in 1444 healthy adolescents [M = 14.39 years (SD = 0.40), 51.5% female] from the European IMAGEN multicenter study. To determine the effects of CP, we applied a two-step approach: (a) testing matched subgroups of low versus high CP, extending into the clinical range [N = 182 per group, M = 14.44 years, (SD = 0.41), 47.3% female] using analysis of variance, and (b) considering (non)linear effects along the CP dimension in the full sample and in the high CP group using multiple regression. We observed no significant cortical or subcortical effect of CP group on brain responses to negative facial affect. In the full sample, regression analyses revealed a significant linear increase of left orbitofrontal cortex (OFC) activity with increasing CP up to the clinical range. In the high CP group, a significant inverted u-shaped effect indicated that left OFC responses decreased again in individuals with high CP. Left OFC activity during negative affective processing which is increasing with CP and decreasing in the highest CP range may reflect on the importance of frontal control mechanisms that counteract the consequences of severe CP by facilitating higher social engagement and better evaluation of social content in adolescents.

Evidence against emotion inference deficits in children with ADHD.

Inconsistent evidence suggests that pediatric attention deficit/hyperactivity disorder (ADHD) may be associated with impairments in the ability to use context clues to infer the emotion states of others. However, the evidence base for these impairments is comprised of data from laboratory-based tests of emotion inference that may be confounded by demands on nonaffective cognitive processes that have been linked with ADHD. The current study builds on our previous study of facial affect recognition to address this limitation and investigate a potential mechanism underlying children's ability to infer emotion state from context clues. To do so, we used a fully crossed, counterbalanced experimental design that systematically manipulated emotion inference and working memory demands in 77 carefully phenotyped children ages 8-13 (Mage = 10.46, SD = 1.54; 66% Caucasian/Non-Hispanic; 42% female) with ADHD (n = 42) and without ADHD (n = 35). Results of Bayesian mixed-model ANOVAs indicated that using context clues to infer the emotion state of others competed for neurocognitive resources with the processes involved in rehearsing/maintaining information within working memory (BF₁₀ = 1.57 × 10¹⁹, d = 0.72). Importantly, there was significant evidence against the critical Group × Condition interaction for response times (BF₀₁ = 4.93), and no significant evidence for this interaction for accuracy (BF₀₁ = 2.40). In other words, children with ADHD do not infer emotions more slowly than children without ADHD (d = 0.13), and their small magnitude impairment in accuracy (d = 0.30) was attributable to their generally less accurate performance on choice-response tasks (i.e., across both emotion and control conditions). Taken together, the evidence indicates that emotion inference abilities are likely unimpaired in pediatric ADHD and that working memory is implicated in the ability to infer emotion from context for all children-not just children with ADHD. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Supplemental material for Of like mind: The (mostly) similar mentalizing of robots and humans.

Mentalizing is the process of inferencing others’ mental states and contributes to an inferential system known as Theory of Mind (ToM)—a system that is critical to human interactions as it facilitates sense-making and prediction of future behaviors. As technological agents like social robots increasingly exhibit hallmarks of intellectual and social agency—and are increasingly integrated into contemporary social life—it is not yet fully understood whether humans hold ToM for such agents. To build on extant research in this domain, five canonical tests that signal implicit mentalizing (white lie detection, intention inferencing, facial affect interpretation, vocal affect interpretation, and false-belief detection) were conducted for an agent (anthropomorphic or machinic robots, or a human) through video-presented (Study 1 <#s6> ) and physically co-present interactions (Study 2 <#s20> ). Findings suggest that mentalizing tendencies for robots and humans are more alike than different, however use of non-literal language, co-present interactivity, and reliance on agent-class heuristics may reduce tendencies to mentalize robots.

Of like mind: The (mostly) similar mentalizing of robots and humans.

Mentalizing is the process of inferencing others’ mental states and contributes to an inferential system known as Theory of Mind (ToM)—a system that is critical to human interactions as it facilitates sense-making and prediction of future behaviors. As technological agents like social robots increasingly exhibit hallmarks of intellectual and social agency—and are increasingly integrated into contemporary social life—it is not yet fully understood whether humans hold ToM for such agents. To build on extant research in this domain, five canonical tests that signal implicit mentalizing (white lie detection, intention inferencing, facial affect interpretation, vocal affect interpretation, and false-belief detection) were conducted for an agent (anthropomorphic or machinic robots, or a human) through video-presented (Study 1 ) and physically co-present interactions (Study 2 ). Findings suggest that mentalizing tendencies for robots and humans are more alike than different, however use of non-literal language, co-present interactivity, and reliance on agent-class heuristics may reduce tendencies to mentalize robots.

Validation of dynamic virtual faces for facial affect recognition.

The ability to recognise facial emotions is essential for successful social interaction. The most common stimuli used when evaluating this ability are photographs. Although these stimuli have proved to be valid, they do not offer the level of realism that virtual humans have achieved. The objective of the present paper is the validation of a new set of dynamic virtual faces (DVFs) that mimic the six basic emotions plus the neutral expression. The faces are prepared to be observed with low and high dynamism, and from front and side views. For this purpose, 204 healthy participants, stratified by gender, age and education level, were recruited for assessing their facial affect recognition with the set of DVFs. The accuracy in responses was compared with the already validated Penn Emotion Recognition Test (ER-40). The results showed that DVFs were as valid as standardised natural faces for accurately recreating human-like facial expressions. The overall accuracy in the identification of emotions was higher for the DVFs (88.25%) than for the ER-40 faces (82.60%). The percentage of hits of each DVF emotion was high, especially for neutral expression and happiness emotion. No statistically significant differences were discovered regarding gender. Nor were significant differences found between younger adults and adults over 60 years. Moreover, there is an increase of hits for avatar faces showing a greater dynamism, as well as front views of the DVFs compared to their profile presentations. DVFs are as valid as standardised natural faces for accurately recreating human-like facial expressions of emotions.

Dysfunctional sleep insufficiency and reduced P3 attentional response to positive social information

One must process others’ facial affect proficiently to facilitate social functioning, and both social functioning and processing of facial affect can be impaired in sleep-deficient states. Based on mood-salience, decreased positive affect and increased negative affect experienced from sleep deficits or from depression might differentially modulate attentional responses to others’ positive and negative facial expressions. The present study examined this possibility by assessing college students’ subjective sleep–wake experience and their P3 event-related potentials responses, which indicate attention allocation. Fifty-five participants completed the Pittsburgh Sleep Quality Index (PSQI), assessments of current and past depressive symptoms, and a modified oddball task including affective facial expression stimuli. Factor analysis was conducted on PSQI data, and structural equation models were estimated to examine associations between PSQI latent factors and oddball P3 amplitude, controlling for current and past depressive symptomatology. Results indicated that one of the two identified PSQI factors (dysfunctional sleep insufficiency) was negatively associated with P3 amplitude in response to happy facial stimuli, while PSQI factor associations with sad and neutral stimuli were non-significant. Findings suggest that particular self-reported sleep–wake problems (dynamic interplay of sleep duration, sleep quality, and psychosocial functioning) can be related to decreased elaborated attention towards others’ positive facial expressions, potentially a marker of decreased inclination towards social rewards. This pattern may be particularly relevant in college, a context that can present important social opportunities and reinforce problematic sleep habits. Future longitudinal research that includes assessments of social functioning and diverse samples would be beneficial.

The Bidirectional View of Mother-Infant Interaction by Gaze and Facial Affect.

The aim of this research was to assess gaze behaviors and facial affect expressions in mother-infant interaction within the frame of self-contingency and interactive contingency, reflecting self-regulation and interactive regulation, respectively. In Model 1, second-by-second changing gaze behaviors (on partner's face/off partner's face) and in Model 2, facial affect expressions (from positive to negative) were examined. Self-contingency reflects the variability or stability in gaze directions and facial affect expressions in each partner. Interactive contingency reflects the degree of mother-infant gaze and facial affect attunement or interactive regulation relative to each other. Sample was composed of 56 healthy mother-infant dyads. All infants were 4 months old, and mean maternal age was 29.61 (SD=3.71). Mother-infant interactions were filmed at the lab. Interactions were coded second-by-second for mother-infant gaze behaviors and facial affect expressions with video microanalysis method. The analysis method was multilevel-multivariate time series analysis. According to Model 1-2, mother-infant gaze behaviors and facial affect expressions were neither too stable nor too variable, rather, the change in gaze behaviors and facial affect expressions in each partner showed predictable patterns. Mothers regulated their gaze behaviors and facial affect expressions in relation to that of their infants. Infants regulated their facial affect expressions in relation to their mothers' facial affect expressions, but infant gaze interactive contingency to mother gaze was marginally significant. In interactions, infants and mothers regulate the rhythms of their own behavior and at the same time contingently coordinate with that of the partner. This bi-directionally regulating environment is the foundation of infants' relationship expectations and bio-socialbehavioral regulation capacity, which may be related to psychopathology in future.

Estimation of continuous valence and arousal levels from faces in naturalistic conditions

Facial affect analysis aims to create new types of human–computer interactions by enabling computers to better understand a person’s emotional state in order to provide ad hoc help and interactions. Since discrete emotional classes (such as anger, happiness, sadness and so on) are not representative of the full spectrum of emotions displayed by humans on a daily basis, psychologists typically rely on dimensional measures, namely valence (how positive the emotional display is) and arousal (how calming or exciting the emotional display looks like). However, while estimating these values from a face is natural for humans, it is extremely difficult for computer-based systems and automatic estimation of valence and arousal in naturalistic conditions is an open problem. Additionally, the subjectivity of these measures makes it hard to obtain good quality data. Here we introduce a novel deep neural network architecture to analyse facial affect in naturalistic conditions with a high level of accuracy. The proposed network integrates face alignment and jointly estimates both categorical and continuous emotions in a single pass, making it suitable for real-time applications. We test our method on three challenging datasets collected in naturalistic conditions and show that our approach outperforms all previous methods. We also discuss caveats regarding the use of this tool, and ethical aspects that must be considered in its application. The annotation of the visual signs of emotions can be important for psychological studies and even human–computer interactions. Instead of only ascribing discrete emotions, Toisoul and colleagues use a single neural network that predicts emotional labels on a spectrum of valence and arousal without separate face-alignment steps.

Fetal Alcohol Exposure Alters BOLD Activation Patterns in Brain Regions Mediating the Interpretation of Facial Affect.

Although deficits in the interpretation of affective facial expressions have been described clinically and in behavioral studies of fetal alcohol spectrum disorders (FASD), effects of prenatal alcohol exposure on the neural networks that mediate affective appraisal have not previously been examined. We administered a nonverbal event-related fMRI affective appraisal paradigm to 64 children (mean age=12.5years; 18 with fetal alcohol syndrome (FAS) or partial FAS (PFAS), 18 nonsyndromal heavily exposed (HE), and 28 controls). Happy, sad, angry, fearful, and neutral faces and pixelated control images were presented sequentially in a randomized order. The child indicated whether the currently displayed face showed the same or different affect as the previous one. Data from whole-brain analyses showed that all groups activated the appropriate face processing neural networks. Region of interest analyses indicated that, compared to HE and control children, the FAS/PFAS group exhibited greater blood oxygenation level-dependent (BOLD) signal changes when processing neutral faces than pixelated images in 2 regions that form part of the visual sensory social brain network, which plays an important role in the initial processing of facial affect. By contrast, BOLD signal when processing angry faces was weaker for the FAS/PFAS group in a region involved in the processing of facial identity and facial expressions and in a region involved in the recognition and selection of behavioral responses to aggressive behavior. These findings of greater BOLD signal in the FAS/PFAS group in response to neutral faces suggest less efficient neural processing of more difficult to interpret emotions, and the weaker BOLD response to angry faces suggests altered processing of angry stimuli. Although behavioral performance did not differ in this relatively simple affective appraisal task, these data suggest that in children with FAS and PFAS, the appraisal of neutral affect and anger is likely to be more effortful in more challenging and dynamic social contexts.

Disentangling 3D/4D Facial Affect Recognition With Faster Multi-View Transformer

In this paper, we propose MiT: a novel multi-view transformer model 1 for 3D/4D facial affect recognition. MiT incorporates patch and position embeddings from various patches of multi-views and uses them for learning various facial muscle movements to showcase an effective recognition performance. We also propose a multi-view loss function that is not only gradient-friendly, and hence speeds up the gradient computation during back-propagation, but it also leverages the correlation associated with the underlying facial patterns among multi-views. Additionally, we offer multi-view weights that are trainable and learnable, and help substantially in training. Finally, we equip our model with distributed performance for faster learning and computational convenience. With the help of extensive experiments, we show that our model outperform the existing methods on widely-used datasets for 3D/4D FER.

White matter abnormalities in misophonia

Misophonia is a condition in which specific ordinary sounds provoke disproportionately strong negative affect and physiological arousal. Evidence for neurobiological abnormalities underlying misophonia is scarce. Since many psychiatric disorders show white matter (WM) abnormalities, we tested for both macro and micro-structural WM differences between misophonia patients and healthy controls.We collected T1-weighted and diffusion-weighted magnetic resonance images from 24 patients and 25 matched controls. We tested for group differences in WM volume using whole-brain voxel-based morphometry and used the significant voxels from this analysis as seeds for probabilistic tractography. After calculation of diffusion tensors, we compared group means for fractional anisotropy, mean diffusivity, and directional diffusivities, and applied tract-based spatial statistics for voxel-wise comparison.Compared to controls, patients had greater left-hemispheric WM volumes in the inferior fronto-occipital fasciculus, anterior thalamic radiation, and body of the corpus callosum connecting bilateral superior frontal gyri. Patients also had lower averaged radial and mean diffusivities and voxel-wise comparison indicated large and widespread clusters of lower mean diffusivity.We found both macro and microstructural WM abnormalities in our misophonia sample, suggesting misophonia symptomatology is associated with WM alterations. These biological alterations may be related to differences in social-emotional processing, particularly recognition of facial affect, and to attention for affective information.

QOL-15. NEURAL NETWORK INTEGRITY FOR FACIAL AFFECT RECOGNITION IN SURVIVORS OF MEDULLOBLASTOMA

BACKGROUNDMedulloblastoma survivors are at risk for social deficits, yet underlying mechanisms are poorly understood.METHODSFacial affect recognition was assessed in 50 medulloblastoma survivors treated with craniospinal radiation (median[range] 21.4[12.5–30.9] years old, 11.0[5.7–22.6] years since diagnosis) and 56 non-cancer age-, sex-, and race-matched controls. Brain activation and connectivity in core regions/nodes of the face perception network (fusiform gyri, occipital gyri, superior temporal sulcus) were examined using structural and functional neuroimaging. Structural networks were constructed from diffusion tensor imaging (DTI) data and individual node strength and efficiency were assessed. Functional MRI (fMRI) was conducted using a 1-back facial affect recognition task with assessment of regional differences in task-related cerebral blood flow (BOLD). Standardized neurocognitive testing was completed with 24 hours of brain imaging.RESULTSMedulloblastoma survivors performed worse on a behavioral measure of facial affect recognition (P=0.003) compared to matched controls. During the facial affect recognition task, controls demonstrated greater BOLD activation of the left and right fusiform gyri and the left and right middle occipital gyri compared to survivors (P’s<0.05, corrected for multiple comparisons). DTI indicated weaker core node strength in survivors in the right lateral occipital gyri (P=0.02) and efficiency was lower in the left (P=0.01) and right (P=0.03) occipital gyri compared to controls.CONCLUSIONSMedulloblastoma survivors have deficits in facial affect recognition and reduced activation and efficiency in brain regions comprising the face perception network compared to matched controls. Interventions targeting this specific skill and neural network may improve social functioning in survivors.

Multimodal affect analysis of psychodynamic play therapy

Objective: We explore state of the art machine learning based tools for automatic facial and linguistic affect analysis to allow easier, faster, and more precise quantification and annotation of children’s verbal and non-verbal affective expressions in psychodynamic child psychotherapy. Method: The sample included 53 Turkish children: 41 with internalizing, externalizing and comorbid problems; 12 in the non-clinical range. We collected audio and video recordings of 148 sessions, which were manually transcribed. Independent raters coded children’s expressions of pleasure, anger, sadness and anxiety using the Children’s Play Therapy Instrument (CPTI). Automatic facial and linguistic affect analysis modalities were adapted, developed, and combined in a system that predicts affect. Statistical regression methods (linear and polynomial regression) and machine learning techniques (deep learning, support vector regression and extreme learning machine) were used for predicting CPTI affect dimensions. Results: Experimental results show significant associations between automated affect predictions and CPTI affect dimensions with small to medium effect sizes. Fusion of facial and linguistic features work best for pleasure predictions; however, for other affect predictions linguistic analyses outperform facial analyses. External validity analyses partially support anger and pleasure predictions. Discussion: The system enables retrieving affective expressions of children, but needs improvement for precision.

Social Cognition Deficits Are Pervasive across Both Classical and Overlap Frontotemporal Dementia Syndromes

Objectives: Frontotemporal dementia (FTD) syndromes are a complex group of disorders characterised by profound changes in behaviour and cognition. Many of the observed behavioural abnormalities are now recognised to be due to impaired social cognition. While deficits in emotion recognition and empathy are well-recognised in behavioural-variant (Bv)FTD, limited information exists about the nature of social cognitive impairment in the language variant primary progressive aphasia (PPA) that includes progressive non-fluent aphasia (PNFA) and semantic dementia (SD), and in the motor variants FTD amyotrophic lateral sclerosis (FTD-ALS) and FTD progressive supranuclear palsy (FTD-PSP). This prospective study sought to explore the nature and profile of social cognition deficits across the spectrum of FTD. Methods: Sixty patients on the FTD spectrum, i.e., classical (16 with BvFTD and 20 with PPA) and overlap FTD syndromes (13 with FTD-ALS and 11 with FTD-PSP) were evaluated by means of the social cognition tasks, the Interpersonal Reactivity Index (IRI) for empathy, and pictures of facial affect (POFA) for emotion recognition. General cognition and behaviour were also assessed. Results: A significant impairment in emotion recognition and empathy was detected in both the classical and overlap FTD syndromes. The recognition of positive emotions was relatively preserved compared to that of negative emotions. Among the FTD subtypes, maximal impairment of empathy was demonstrated in FTD-PSP. Conclusion: Social cognition impairment is pervasive across the spectrum of FTD disorders, and tests of emotion recognition and empathy are clinically useful to identify the nature of behavioural problems in both classical and overlap FTD. Our findings also have implications for understanding the neural basis of social cognition in FTD.

A Distinct Neurophenotype of Fearful Face Processing in Alcohol Use Disorder With and Without Comorbid Anxiety.

Individuals with alcohol use disorder (AUD) can present with comorbid anxiety symptoms and often have deficits in emotional processing. Previous research suggests brain response is altered during facial affect recognition tasks, especially in limbic areas, due to either AUD or anxiety symptomology; however, the impact of both AUD and clinically significant anxiety symptoms during these tasks has not yet been examined. In this study, we investigated neural activation differences during an emotional face-matching task. Participants (N=232) underwent fMRI scanning, as part of a larger study. Three groups were investigated: individuals with diagnosed AUD and elevated anxiety traits (AUD+ANX, n=90), individuals with diagnosed AUD but non-clinically significant levels of anxiety (AUD-ANX, n=39), and healthy controls (HC, n=103). Our results illustrate distinct neurophenotypes of AUD, where individuals with comorbid anxiety symptomology have blunted emotional face processing while those with singular AUD are hyperresponsive. This suggests AUD with anxiety symptomology may have a unique neurobiological underpinning, and treatment and intervention should be tailored to individual constellations of symptoms.

Does Social Inferencing Differ in Men and Women after TBI?

Social inferencing skills are often compromised after a traumatic brain injury (TBI). Some studies suggest that women with TBI are less impaired than their male counterparts at facial affect recognition. The study objective was to examine sex differences in additional aspects of social inferencing in males and females with and without a TBI.

Multimodal emotion perception: Influences of autism spectrum disorder and autism-like traits

Individuals with autism spectrum disorder may exhibit impaired production and perception of facial and vocal affect cues. For example, both neurotypical and autistic adults tend to interpret affect less accurately from autistic faces and voices, with autistic listeners displaying overall lower identification accuracy. However, earlier studies compared neurotypical and autistic emotion perception under exclusively unimodal conditions (facial or verbal cues alone) or exclusively multimodal conditions in which utterances and static faces were presented sequentially. This study assesses neurotypical adults' accuracy and response time in a forced-choice emotion identification task with video-only, audio-only, and audiovisual recordings of neurotypical and autistic talkers. The study also investigates the relation between listeners' task performance and Autism-Spectrum Quotient (AQ) scores (degree of autism-like traits). Preliminary results show listeners are less accurate when identifying autistic emotional expressions, compared to neurotypical expressions, while talker type has no main effect on response time. Results also suggest a relation between AQ score and emotion-perception ability (accuracy and response time) in neurotypical respondents. This research will provide insight into emotion perception and production differences, along with implications for interpersonal functioning, in autism and under the broader autism phenotype within the neurotypical population. [Work supported by the Indiana University Hutton Honors College.]

Deficits in recognizing female facial expressions related to social network in cocaine-addicted men

BackgroundThe ability to accurately recognise facial expressions of emotion is crucial for social functioning and maintaining healthy relationships. Recognising the emotional state of others allows us to respond to their needs and adjust our behaviour appropriately. Impairments in facial affect recognition have been reported in chronic cocaine users but little is known whether these contribute to their difficulties in social situations. MethodsWe assessed facial emotional expression recognition in forty-five men with cocaine use disorder (CUD) and forty-four healthy control participants. Using standardised questionnaires, we also collected information on perceived social support, social provision and community integration. ResultsOur results found that male cocaine users had greater difficulty in recognising female emotional facial expressions than male controls. This effect was not explained by demographic variables but it was associated with their social network; including social support, social provisions and community integration. ConclusionOur findings suggest that men with CUD have greater difficulty in identifying emotional expression in female faces, which is linked with their social support networks. This may play an important role in misunderstanding non-verbal communications that contribute to destabilising friendship and family ties typically seen in drug addiction. Addressing deficits in recognising female emotional expressions may be an important piece of information for counselling and other interventions.

Using mobile eye-tracking technology to examine adolescent daughters' attention to maternal affect during a conflict discussion.

Attention to socio-emotional stimuli (i.e., affect-biased attention) is an integral component of emotion regulation and human communication. Given the strong link between maternal affect and adolescent behavior, maternal affect may be a critical influence on adolescent affect-biased attention during mother-child interaction. However, prior methodological constraints have precluded fine-grained examinations of factors such as maternal affect on adolescent attention during real-world social interaction. Therefore, this pilot study capitalized on previously validated technological advances by using mobile eye-tracking and facial affect coding software to quantify the influence of maternal affect on adolescents' attention to the mother during a conflict discussion. Results from 7,500 to 9,000 time points sampled for each mother-daughter dyad (n=28) indicated that both negative and positive maternal affect, relative to neutral, elicited more adolescent attentional avoidance of the mother (ORs=2.68-9.20), suggesting that typically developing adolescents may seek to avoid focusing on maternal affect of either valence during a conflict discussion. By examining the moment-to-moment association between in vivo displays of maternal affect and subsequent adolescent attention toward the mother's face, these results provide preliminary evidence that maternal affect moderates adolescent attention. Our findings are consistent with cross-species approach-avoidance models suggesting that offspring respond to affectively charged conversations with greater behavioral avoidance or deference.