Using computer-vision and machine learning to automate facial coding of positive and negative affect intensity

Facial expressions are fundamental to interpersonal communication, including social interaction, and allow people of different ages, cultures, and languages to quickly and reliably convey emotional information. Historically, facial expression research has followed from discrete emotion theories, which posit a limited number of distinct affective states that are represented with specific patterns of facial action. Much less work has focused on dimensional features of emotion, particularly positive and negative affect intensity. This is likely, in part, because achieving inter-rater reliability for facial action and affect intensity ratings is painstaking and labor-intensive. We use computer-vision and machine learning (CVML) to identify patterns of facial actions in 4,648 video recordings of 125 human participants, which show strong correspondences to positive and negative affect intensity ratings obtained from highly trained coders. Our results show that CVML can both (1) determine the importance of different facial actions that human coders use to derive positive and negative affective ratings when combined with interpretable machine learning methods, and (2) efficiently automate positive and negative affect intensity coding on large facial expression databases. Further, we show that CVML can be applied to individual human judges to infer which facial actions they use to generate perceptual emotion ratings from facial expressions.

Early caregiving stress exposure moderates the relation between respiratory sinus arrhythmia reactivity at 1 month and biobehavioral outcomes at age 3

There is a growing scientific interest in the psychophysiological functioning of children living in low-socioeconomic status (SES) contexts, though this research is complicated by knowledge that physiology-behavior relations often operate differently in these environments among adults. Importantly, such research is made more difficult because SES may be a proxy for a wide range of risk factors including poor caregiving and exposure to parental substance use. We used factor analysis to organize risk-exposure data collected from 827 children-many of whom were raised in low-SES contexts and exposed to substances prenatally-into dissociable components including economic stress, caregiving stress (e.g., stress the caregiver may experience, including parental psychopathology), and postnatal substance exposure. These factors, along with respiratory sinus arrhythmia (RSA) reactivity at age 1 month, were used to predict behavior dysregulation and resting RSA at age 3 years. A significant RSA Reactivity × Caregiving Stress interaction indicated that infants who exhibited high RSA reactivity at 1 month experienced the greatest behavior dysregulation at 3 years, but only when they were exposed to high levels of caregiving stress. Among African Americans, the highest resting RSA at 3 years was found in infants with less RSA reactivity, but only if they also experienced less caregiving stress. Our work is consistent with biological sensitivity to context, adaptive calibration, and allostatic load models, and highlights the importance of studying Physiology × Environment interactions in low-SES contexts for predicting behavior and resting RSA.

Evidence for latent classes of IQ in young children with autism spectrum disorder

Autism is currently viewed as a spectrum condition that includes strikingly different severity levels; IQ is consistently described as one of the primary aspects of the heterogeneity in autism. To investigate the possibility of more than one distinct subtype of autism based on IQ both latent class analysis and taxometrics methods were used to classify Mullen IQs in a sample of 456 children with autism spectrum disorder. We found evidence for multiple IQbased subgroups using both methods. Groups differed in level of intellectual functioning and patterns of verbal versus nonverbal ability. Results support the notion of distinct subtypes of autism that differ in severity of intellectual ability, patterns of cognitive strengths and weaknesses, and severity of autism symptoms.

Vagal tone, development, and Gray's motivational theory: toward an integrated model of autonomic nervous system functioning in psychopathology

In the last decade, cardiac vagal tone has emerged as a psychophysiological marker of many aspects of behavioral functioning in both children and adults. Research efforts during this time have produced an extensive list of vagal tone correlates that includes temperamental variables as well as both anxious/internalizing and disruptive/externalizing behaviors. This potentially confusing state of affairs is compounded by developmental shifts in vagal tone behavior relations that to date have not been elucidated. In this paper, the vagal tone literature is reviewed, and discrepancies, including the lack of specificity of vagal tone as a psychophysiological marker, are clarified. Such clarification requires that we (a) view vagal tone-behavior relations in developmental context, (b) juxtapose vagal tone-behavior relations in typical and atypical samples, and (c) consider the parasympathetic underpinnings of vagal tone as but one component in a broader model of autonomic nervous system functioning. Such a model is provided by combining Gray's motivational theory with Porges's polyvagal theory. Together these models account for behavioral and emotional differences in a diverse range of psychological disorders that are not differentiated by either model alone. Moreover, use of the integrated model offers a theory-driven approach to the study of autonomic nervous system-behavior relations.