Beyond Likert ratings: Improving the robustness of developmental research measurement using best-worst scaling.

Some of the 'best practice' approaches to ensuring reproducibility of research can be difficult to implement in the developmental and clinical domains, where sample sizes and session lengths are constrained by the practicalities of recruitment and testing. For this reason, an important area of improvement to target is the reliability of measurement. Here we demonstrate that best-worst scaling (BWS) provides a superior alternative to Likert ratings for measuring children's subjective impressions. Seventy-three children aged 5-6 years rated the trustworthiness of faces using either Likert ratings or BWS over two sessions. Individual children's ratings in the BWS condition were significantly more consistent from session 1 to session 2 than those in the Likert condition, a finding we also replicate with a large adult sample (N = 72). BWS also produced more reliable ratings at the group level than Likert ratings in the child sample. These findings indicate that BWS is a developmentally appropriate response format that can deliver substantial improvements in reliability of measurement, which can increase our confidence in the robustness of findings with children.

Individual differences in trust evaluations are shaped mostly by environments, not genes.

People evaluate a stranger's trustworthiness from their facial features in a fraction of a second, despite common advice "not to judge a book by its cover." Evaluations of trustworthiness have critical and widespread social impact, predicting financial lending, mate selection, and even criminal justice outcomes. Consequently, understanding how people perceive trustworthiness from faces has been a major focus of scientific inquiry, and detailed models explain how consensus impressions of trustworthiness are driven by facial attributes. However, facial impression models do not consider variation between observers. Here, we develop a sensitive test of trustworthiness evaluation and use it to document substantial, stable individual differences in trustworthiness impressions. Via a twin study, we show that these individual differences are largely shaped by variation in personal experience, rather than genes or shared environments. Finally, using multivariate twin modeling, we show that variation in trustworthiness evaluation is specific, dissociating from other key facial evaluations of dominance and attractiveness. Our finding that variation in facial trustworthiness evaluation is driven mostly by personal experience represents a rare example of a core social perceptual capacity being predominantly shaped by a person's unique environment. Notably, it stands in sharp contrast to variation in facial recognition ability, which is driven mostly by genes. Our study provides insights into the development of the social brain, offers a different perspective on disagreement in trust in wider society, and motivates new research into the origins and potential malleability of face evaluation, a critical aspect of human social cognition.

Do facial first impressions reflect a shared social reality?

Influential facial impression models have repeatedly shown that trustworthiness, youthful-attractiveness, and dominance dimensions subserve a wide variety of first impressions formed from strangers' faces, suggestive of a shared social reality. However, these models are built from impressions aggregated across observers. Critically, recent work has now shown substantial inter-observer differences in facial impressions, raising the important question of whether these dimensional models based on aggregated group data are meaningful at the individual observer level. We addressed this question with a novel case series approach, using factor analyses of ratings of twelve different traits to build individual models of facial impressions for different observers. Strikingly, three dimensions of trustworthiness, youthful/attractiveness, and competence/dominance appeared across the majority of these individual observer models, demonstrating that the dimensional approach is indeed meaningful at the individual level. Nonetheless, we also found differences in the stability of the competence/dominance dimension across observers. Taken together, results suggest that individual differences in impressions arise in the context of a largely common structure that supports a shared social reality.

Are expression aftereffects fully explained by tilt adaptation?

Facial expressions are used as critical social cues in everyday life. Adaptation to expressions causes expression aftereffects. These aftereffects are thought to reflect the operation of face-selective neural mechanisms, and are used by researchers to investigate the nature of those mechanisms. However, recent evidence suggests that expression aftereffects could be at least partially explained by the inheritance of lower-level tilt adaptation through the visual hierarchy. We investigated whether expression aftereffects could be entirely explained by tilt adaptation. Participants completed an expression adaptation task in which we controlled for the influence of tilt by changing the orientation of the adaptor relative to the test stimuli. Although tilt adaptation appeared to make some contribution to the expression aftereffect, robust expression aftereffects still remained after minimizing tilt inheritance, indicating that expression aftereffects cannot be fully explained by tilt adaptation. There was also significant reduction in the expression aftereffects after inverting the adapting face, providing evidence that face-selective processing is involved in these aftereffects.

Best-worst scaling improves measurement of first impressions.

A common goal in psychological research is the measurement of subjective impressions, such as first impressions of faces. These impressions are commonly measured using Likert ratings. Although these ratings are simple to administer, they are associated with response issues that can limit reliability. Here we examine best-worst scaling (BWS), a forced-choice method, as a potential alternative to Likert ratings for measuring participants' facial first impressions. We find that at the group level, BWS scores correlated almost perfectly with Likert scores, indicating that the two methods measure the same impressions. However, at the individual participant level BWS outperforms Likert ratings, both in terms of ability to predict preferences in a third task, and in terms of test-retest reliability. These benefits highlight the power of BWS, particularly for use in individual differences research.

Expression Recognition Difficulty Is Associated with Social But Not Attention-to-Detail Autistic Traits and Reflects Both Alexithymia and Perceptual Difficulty.

Autistic people often show difficulty with facial expression recognition. However, the degree of difficulty varies widely, which might reflect varying symptom profiles. We examined three domains of autistic traits in the typical population and found that more autistic-like social skills were associated with greater difficulty labelling expressions, and more autistic-like communication was associated with greater difficulty labelling and perceptually discriminating between expressions. There were no associations with autistic-like attention to detail. We also found that labelling, but not perceptual, difficulty was mediated by alexithymia. We found no evidence that labelling or perceptual difficulty was mediated by weakened adaptive coding. Results suggest expression recognition varies between the sub-clinical expressions of autistic symptom domains and reflects both co-occurring alexithymia and perceptual difficulty.

Perceptual experience shapes our ability to categorize faces by national origin: A new other-race effect.

People are better at recognizing own-race than other-race faces. This other-race effect has been argued to be the result of perceptual expertise, whereby face-specific perceptual mechanisms are tuned through experience. We designed new tasks to determine whether other-race effects extend to categorizing faces by national origin. We began by selecting sets of face stimuli for these tasks that are typical in appearance for each of six nations (three Caucasian, three Asian) according to people from those nations (Study 1). Caucasian and Asian participants then categorized these faces by national origin (Study 2). Own-race faces were categorized more accurately than other-race faces. In contrast, Asian American participants, with more extensive other-race experience than the first Asian group, categorized other-race faces better than own-race faces, demonstrating a reversal of the other-race effect. Therefore, other-race effects extend to the ability to categorize faces by national origin, but only if participants have greater perceptual experience with own-race, than other-race faces. Study 3 ruled out non-perceptual accounts by showing that Caucasian and Asian faces were sorted more accurately by own-race than other-race participants, even in a sorting task without any explicit labelling required. Together, our results demonstrate a new other-race effect in sensitivity to national origin of faces that is linked to perceptual expertise.

Beyond opponent coding of facial identity: Evidence for an additional channel tuned to the average face.

Face identity can be represented in a multidimensional space centered on the average. It has been argued that the average acts as a perceptual norm, with the norm coded implicitly by balanced activation in pairs of channels that respond to opposite extremes of face dimensions (two-channel model). In Experiment 1 we used face identity aftereffects to distinguish this model from a narrow-band multichannel model with no norm. We show that as adaptors become more extreme, aftereffects initially increase sharply and then plateau. Crucially there is no decrease, ruling out narrow-band multichannel coding, but consistent with a two-channel norm-based model. However, these results leave open the possibility that there may be a third channel, tuned explicitly to the norm (three-channel model). In Experiment 2 we show that alternating adaptation widens the range identified as the average whereas adaptation to the average narrows the range, consistent with the three-channel model. Explicit modeling confirmed the three-channel model as the best fit for the combined data from both experiments. However, a two-channel model with decision criteria allowed to vary between adapting conditions, also provided a very good fit. These results support opponent, norm-based coding of face identity with additional explicit coding of the norm. (PsycINFO Database Record

Adaptive face coding contributes to individual differences in facial expression recognition independently of affective factors.

There are large, reliable individual differences in the recognition of facial expressions of emotion across the general population. The sources of this variation are not yet known. We investigated the contribution of a key face perception mechanism, adaptive coding, which calibrates perception to optimize discrimination within the current perceptual "diet." We expected that a facial expression system that readily recalibrates might boost sensitivity to variation among facial expressions, thereby enhancing recognition ability. We measured adaptive coding strength with an established facial expression aftereffect task and measured facial expression recognition ability with 3 tasks optimized for the assessment of individual differences. As expected, expression recognition ability was positively associated with the strength of facial expression aftereffects. We also asked whether individual variation in affective factors might contribute to expression recognition ability, given that clinical levels of such traits have previously been linked to ability. Expression recognition ability was negatively associated with self-reported anxiety but not with depression, mood, or degree of autism-like or empathetic traits. Finally, we showed that the perceptual factor of adaptive coding contributes to variation in expression recognition ability independently of affective factors. (PsycINFO Database Record

Facial expression coding in children and adolescents with autism: Reduced adaptability but intact norm-based coding.

Individuals with autism spectrum disorder (ASD) can have difficulty recognizing emotional expressions. Here, we asked whether the underlying perceptual coding of expression is disrupted. Typical individuals code expression relative to a perceptual (average) norm that is continuously updated by experience. This adaptability of face-coding mechanisms has been linked to performance on various face tasks. We used an adaptation aftereffect paradigm to characterize expression coding in children and adolescents with autism. We asked whether face expression coding is less adaptable in autism and whether there is any fundamental disruption of norm-based coding. If expression coding is norm-based, then the face aftereffects should increase with adaptor expression strength (distance from the average expression). We observed this pattern in both autistic and typically developing participants, suggesting that norm-based coding is fundamentally intact in autism. Critically, however, expression aftereffects were reduced in the autism group, indicating that expression-coding mechanisms are less readily tuned by experience. Reduced adaptability has also been reported for coding of face identity and gaze direction. Thus, there appears to be a pervasive lack of adaptability in face-coding mechanisms in autism, which could contribute to face processing and broader social difficulties in the disorder.

A new other-race effect for gaze perception.

The well-known other-race effect in face recognition has been widely studied, both for its theoretical insights into the nature of face expertise and because of its social and forensic importance. Here we demonstrate an other-race effect for the perception of a simple visual signal provided by the eyes, namely gaze direction. In Study 1, Caucasian and Asian participants living in Australia both showed greater perceptual sensitivity to detect direct gaze in own-race than other-race faces. In Study 2, Asian (Chinese) participants living in Australia and Asian (Chinese) participants living in Hong Kong both showed this other-race effect, but Caucasian participants did not. Despite this inconsistency, meta-analysis revealed a significant other-race effect when results for all 5 participant groups from corresponding conditions in the 2 studies were combined. These results demonstrate a new other-race effect for the perception of the simple, but socially potent, cue of direct gaze. When identical morphed-race eyes were inserted into the faces, removing race-specific eye cues, no other-race effect was found (with 1 exception). Thus, the balance of evidence implicated perceptual expertise, rather than social motivation, in the other-race effect for detecting direct gaze. (PsycINFO Database Record

Aftereffects support opponent coding of expression.

We used aftereffects to investigate the coding mechanisms underlying perception of facial expression. Recent evidence that some dimensions are common to the coding of both expression and identity suggests that the same type of coding system could be used for both attributes. Identity is adaptively opponent coded by pairs of neural populations tuned to opposite extremes of relevant dimensions. Therefore, the authors hypothesized that expression would also be opponent coded. An important line of support for opponent coding is that aftereffects increase with adaptor extremity (distance from an average test face) over the full natural range of possible faces. Previous studies have reported that expression aftereffects increase with adaptor extremity. Critically, however, they did not establish the extent of the natural range and so have not ruled out a decrease within that range that could indicate narrowband, multichannel coding. Here the authors show that expression aftereffects, like identity aftereffects, increase linearly over the full natural range of possible faces and remain high even for impossibly distorted adaptors. These results suggest that facial expression, like face identity, is opponent coded. (PsycINFO Database Record

The timecourse of expression aftereffects.

Adaptation to facial expressions produces aftereffects that bias perception of subsequent expressions away from the adaptor. Studying the temporal dynamics of an aftereffect can help us to understand the neural processes that underlie perception, and how they change with experience. Little is known about the temporal dynamics of the expression aftereffect. We conducted two experiments to measure the timecourse of this aftereffect. In Experiment 1 we examined how the size of the aftereffect varies with changes in the duration of the adaptor and test stimuli. We found that the expression aftereffect follows the classic timecourse pattern of logarithmic build-up and exponential decay that has been demonstrated for many lower level aftereffects, as well as for facial identity and figural face aftereffects. This classic timecourse pattern suggests that the adaptive calibration mechanisms of facial expression are similar to those of lower level visual stimuli, and is consistent with a perceptual locus for the adaptation aftereffect. We also found that aftereffects could be generated by as little as 1 s of adaptation, and in some conditions lasted for as long as 3200 ms. We extended this last finding in Experiment 2, exploring the longevity of the expression aftereffect by adding a stimulus-free gap of varying duration between adaptation and test. We found that significant expression aftereffects were still present 32 s after adaptation. The persistence of the expression aftereffect suggests that they may have a considerable impact on day-to-day expression perception.

How distinct is the coding of face identity and expression? Evidence for some common dimensions in face space.

Traditional models of face perception emphasize distinct routes for processing face identity and expression. These models have been highly influential in guiding neural and behavioural research on the mechanisms of face perception. However, it is becoming clear that specialised brain areas for coding identity and expression may respond to both attributes and that identity and expression perception can interact. Here we use perceptual aftereffects to demonstrate the existence of dimensions in perceptual face space that code both identity and expression, further challenging the traditional view. Specifically, we find a significant positive association between face identity aftereffects and expression aftereffects, which dissociates from other face (gaze) and non-face (tilt) aftereffects. Importantly, individual variation in the adaptive calibration of these common dimensions significantly predicts ability to recognize both identity and expression. These results highlight the role of common dimensions in our ability to recognize identity and expression, and show why the high-level visual processing of these attributes is not entirely distinct.

How is facial expression coded?

Facial expression is theorized to be visually represented in a multidimensional expression space, relative to a norm. This norm-based coding is typically argued to be implemented by a two-pool opponent coding system. However, the evidence supporting the opponent coding of expression cannot rule out the presence of a third channel tuned to the center of each coded dimension. Here we used a paradigm not previously applied to facial expression to determine whether a central-channel model is necessary to explain expression coding. Participants identified expressions taken from a fear/antifear trajectory, first at baseline and then in two adaptation conditions. In one condition, participants adapted to the expression at the center of the trajectory. In the other condition, participants adapted to alternating images from the two ends of the trajectory. The range of expressions that participants perceived as lying at the center of the trajectory narrowed in both conditions, a pattern that is not predicted by the central-channel model but can be explained by the opponent-coding model. Adaptation to the center of the trajectory also increased identification of both fear and antifear, which may indicate a functional benefit for adaptive coding of facial expression.

Facial trustworthiness judgments in children with ASD are modulated by happy and angry emotional cues.

Appearance-based trustworthiness inferences may reflect the misinterpretation of emotional expression cues. Children and adults typically perceive faces that look happy to be relatively trustworthy and those that look angry to be relatively untrustworthy. Given reports of atypical expression perception in children with Autism Spectrum Disorder (ASD), the current study aimed to determine whether the modulation of trustworthiness judgments by emotional expression cues in children with ASD is also atypical. Cognitively-able children with and without ASD, aged 6-12 years, rated the trustworthiness of faces showing happy, angry and neutral expressions. Trust judgments in children with ASD were significantly modulated by overt happy and angry expressions, like those of typically-developing children. Furthermore, subtle emotion cues in neutral faces also influenced trust ratings of the children in both groups. These findings support a powerful influence of emotion cues on perceived trustworthiness, which even extends to children with social cognitive impairments.

Nine-year-old children use norm-based coding to visually represent facial expression.

Children are less skilled than adults at making judgments about facial expression. This could be because they have not yet developed adult-like mechanisms for visually representing faces. Adults are thought to represent faces in a multidimensional face-space, and have been shown to code the expression of a face relative to the norm or average face in face-space. Norm-based coding is economical and adaptive, and may be what makes adults more sensitive to facial expression than children. This study investigated the coding system that children use to represent facial expression. An adaptation aftereffect paradigm was used to test 24 adults and 18 children (9 years 2 months to 9 years 11 months old). Participants adapted to weak and strong antiexpressions. They then judged the expression of an average expression. Adaptation created aftereffects that made the test face look like the expression opposite that of the adaptor. Consistent with the predictions of norm-based but not exemplar-based coding, aftereffects were larger for strong than weak adaptors for both age groups. Results indicate that, like adults, children's coding of facial expressions is norm-based.