Uncovering the semantics of concepts using GPT-4.

The ability of recent Large Language Models (LLMs) such as GPT-3.5 and GPT-4 to generate human-like texts suggests that social scientists could use these LLMs to construct measures of semantic similarity that match human judgment. In this article, we provide an empirical test of this intuition. We use GPT-4 to construct a measure of typicality-the similarity of a text document to a concept. We evaluate its performance against other model-based typicality measures in terms of the correlation with human typicality ratings. We conduct this comparative analysis in two domains: the typicality of books in literary genres (using an existing dataset of book descriptions) and the typicality of tweets authored by US Congress members in the Democratic and Republican parties (using a novel dataset). The typicality measure produced with GPT-4 meets or exceeds the performance of the previous state-of-the art typicality measure we introduced in a recent paper [G. Le Mens, B. Kovács, M. T. Hannan, G. Pros Rius, Sociol. Sci. 2023, 82-117 (2023)]. It accomplishes this without any training with the research data (it is zero-shot learning). This is a breakthrough because the previous state-of-the-art measure required fine-tuning an LLM on hundreds of thousands of text documents to achieve its performance.

Sexual Dimorphism: The Interrelation of Shape and Color.

Sex-typicality displayed as sexual dimorphism of the human face is a key feature enabling sex recognition. It is also believed to be a cue for perceiving biological quality and it plays an important role in the perception of attractiveness. Sexual dimorphism of human faces has two main components: sexual shape dimorphism of various facial features and sexual color dimorphism, generally manifested as dimorphism of skin luminance, where men tend to be darker than women. However, very little is known about the mutual relationship of these two facets. We explored the interconnection between the dimorphism of face shape and dimorphism of face color in three visually distinct populations (Cameroonian, Czech, and Vietnamese). Our results indicated that populations which showed a significant dimorphism in skin luminance (Cameroon, Vietnam) had low levels of sexual shape dimorphism, while a population with higher levels of sexual shape dimorphism (Czech Republic) did not exhibit a significant dimorphism of skin luminance. These findings suggest a possible compensatory mechanism between various domains of sexual dimorphism in populations differing in the levels of shape and color dimorphism.

Pornography Use and the Acceptance of Gender Norm Violation in a School Context.

Although media effect studies have quite extensively investigated the association between pornography use and gendered attitudes, some questions remain. The present study aimed to address two of these questions by exploring how gendered attitudes and gender beliefs may be influenced by gender typicality and pornography use. First, the literature has not yet accounted for individual differences based on gender typicality. Second, the influence of pornography use on gender beliefs going beyond pornography's script application is understudied. This online cross-sectional study (N = 1,440, Mage = 23.86, SD = 4.79) contributes to the field by investigating the indirect association between pornography use and acceptance of gender norm violation through gendered attitudes and the moderating role of gender typicality. Acceptance of gender norm violation was measured via vignettes describing a school context in which a teacher and a student violated gender norms. Findings indicated that gendered attitudes negatively relate to the acceptance of gender norm violation. Moreover, compared to women, men's pornography use indirectly relates to lower acceptance rates through gendered attitudes. Additionally, for men, specific levels of gender typicality and atypicality form a strengthening and buffering role, respectively. This applies to the association between pornography use and gendered attitudes as well as to the indirect relationship of pornography use with acceptance of gender norm violation. These findings suggest that pornography use may also affect gender beliefs that are unrelated to the scripts present in pornography. Future studies should take into account the type of preferred pornography and unravel the specific impact of women's pornography use.

Essential shape: The role of essentialist beliefs in preferences for misshapen produce.

Understanding the reasons behind consumer rejection of misshapen produce is important because of its dramatic consequences on food waste. Drawing on psychological essentialism, we conduct seven studies to investigate the role of produce shape-related essentialist beliefs (i.e., the belief that produce shape is determined by an underlying and unobservable essence) in consumers' evaluations of misshapen produce. Our findings show that essentialist beliefs about produce shape are divided into four distinct dimensions (i.e., biological basis, discreteness, informativeness, and immutability). We find that endorsing discreteness beliefs decreases preferences for misshapen produce. Immutability beliefs produce the opposite effect, whereas informativeness and biological basis have little to no effect. These effects operate through typicality judgments. These findings suggest that considering essentialist beliefs have the potential to help enhance the acceptance of edible albeit misshapen produce.

Entropy of the Canonical Occupancy (Macro) State in the Quantum Measurement Theory.

The paper analyzes the probability distribution of the occupancy numbers and the entropy of a system at the equilibrium composed by an arbitrary number of non-interacting bosons. The probability distribution is obtained through two approaches: one involves tracing out the environment from a bosonic eigenstate of the combined environment and system of interest (the empirical approach), while the other involves tracing out the environment from the mixed state of the combined environment and system of interest (the Bayesian approach). In the thermodynamic limit, the two coincide and are equal to the multinomial distribution. Furthermore, the paper proposes to identify the physical entropy of the bosonic system with the Shannon entropy of the occupancy numbers, fixing certain contradictions that arise in the classical analysis of thermodynamic entropy. Finally, by leveraging an information-theoretic inequality between the entropy of the multinomial distribution and the entropy of the multivariate hypergeometric distribution, Bayesianism of information theory and empiricism of statistical mechanics are integrated into a common "infomechanical" framework.

Facial typicality and attractiveness reflect an ideal dimension of face structure.

Face perception and recognition are important processes for social interaction and communication among humans, so understanding how faces are mentally represented and processed has major implications. At the same time, faces are just some of the many stimuli that we encounter in our everyday lives. Therefore, more general theories of how we represent objects might also apply to faces. Contemporary research on the mental representation of faces has centered on two competing theoretical frameworks that arose from more general categorization research: prototype-based face representation and exemplar-based face representation. Empirically distinguishing between these frameworks is difficult and neither one has been ruled out. In this paper, we advance this area of research in three ways. First, we introduce two additional frameworks for mental representation of categories, varying abstraction and ideal representation, which have not been applied to face perception and recognition before. Second, we fit formal computational models of all four of these theories to human perceptual judgments of the typicality and attractiveness (a strong correlate of typicality) of 100 young adult Caucasian female faces, with the models expressed within a face space derived from facial similarity judgments via multidimensional scaling. Third, we predict the perceived typicality and attractiveness of the faces using these models and compare the predictive performance of each to the empirical data. We found that of all four models, the ideal representation model provided the best account of perceived typicality and attractiveness for the present set of faces, although all models showed discrepancies from the empirical data. These findings demonstrate the relevance of mental categorization processes for representing faces.

From likely to likable: The role of statistical typicality in human social assessment of faces.

Humans readily form social impressions, such as attractiveness and trustworthiness, from a stranger's facial features. Understanding the provenance of these impressions has clear scientific importance and societal implications. Motivated by the efficient coding hypothesis of brain representation, as well as Claude Shannon's theoretical result that maximally efficient representational systems assign shorter codes to statistically more typical data (quantified as log likelihood), we suggest that social "liking" of faces increases with statistical typicality. Combining human behavioral data and computational modeling, we show that perceived attractiveness, trustworthiness, dominance, and valence of a face image linearly increase with its statistical typicality (log likelihood). We also show that statistical typicality can at least partially explain the role of symmetry in attractiveness perception. Additionally, by assuming that the brain focuses on a task-relevant subset of facial features and assessing log likelihood of a face using those features, our model can explain the "ugliness-in-averageness" effect found in social psychology, whereby otherwise attractive, intercategory faces diminish in attractiveness during a categorization task.

A new facet of category theory: cognitive disadvantage and its implications for safety in the cockpit.

The typicality effect suggests typical category members provide a cognitive advantage, such as being quicker and easier to recognise and describe. The reverse effect has not been explored in an applied environment. Non-typical flight safety events appear to pose problems for pilots, leading to delayed recognition and ineffective use of checklists. Fifty-six airline pilots completed an experiment that tested a real-world typicality gradient, comparing pilot performance on a group of four non-typical events against four randomly selected events. Non-typical flight safety events elicited a greater number of response errors and a greater response latency when compared with a random selection of safety events. We specify and measure cognitive disadvantage and suggest innovations in pilot education, such as locating troublesome events and improving recognition guidance. Our new findings can be used to better prepare pilots for event diversity and inform safety in other work systems of interest to ergonomics. Practitioner summary: Typical safety events in work environments provide a cognitive dividend, supporting effective recognition and response. In this study, we frame and measure the opposite effect, the cognitive disadvantages of non-typical events. Non-typical events pose significant risk in work systems such as air transport, and we suggest innovations in pilot knowledge and training that make use of this approach.

The impact of ChatGPT on human data collection: A case study involving typicality norming data.

Tools like ChatGPT, which allow people to unlock the potential of large language models (LLMs), have taken the world by storm. ChatGPT's ability to produce written output of remarkable quality has inspired, or forced, academics to consider its consequences for both research and education. In particular, the question of what constitutes authorship, and how to evaluate (scientific) contributions has received a lot of attention. However, its impact on (online) human data collection has mostly flown under the radar. The current paper examines how ChatGPT can be (mis)used in the context of generating norming data. We found that ChatGPT is able to produce sensible output, resembling that of human participants, for a typicality rating task. Moreover, the test-retest reliability of ChatGPT's ratings was similar to that of human participants tested 1 day apart. We discuss the relevance of these findings in the context of (online) human data collection, focusing both on opportunities (e.g., (risk-)free pilot data) and challenges (e.g., data fabrication).

Typicality modulates the visual awareness of objects.

In the present study, we assessed whether typicality can influence the visual awareness of objects. Participants tracked moving images of objects and counted how often members of one category bounced off the edges of the display. On the last trial, an unexpected object moved across the display. In our first two experiments, this object could belong to the same category as the tracked or untracked objects. While participants were more likely to notice atypical members of the untracked category, this pattern of results reversed when participants tracked atypical objects. In our last two experiments, the unexpected object could belong to the same category as the tracked objects or a new category of objects. In this case, participants were more likely to notice typical members of both the tracked category and the new category. Together, these findings suggest that typicality can modulate the visual awareness of objects.

Gender Typicality, Pressure to Conform to Gender Norms, and Anti-Fat and Appearance Stereotypes in Girls.

This mixed-methods study investigated the relation among gender identity (i.e., self-perceived gender similarity to girls and boys; self-perceived parental and peer pressure to conform to gender norms) and stereotyping about weight (i.e., anti-fat stereotypes), weight change, and appearance in 83 girls ages 6-9 (Mage = 7.60 years, SD = .85; 65% White, 16% Mixed/Other, 11% Black, 8% Latina) in the U.S. Stereotypes about weight change were assessed with open-ended responses (i.e., qualitatively), and the rest of the constructs were assessed with closed-ended responses (i.e., quantitatively). There was a positive association between pressure from parents to conform to gender norms and appearance stereotypes, and between pressure from peers and negative stereotypes about the fat and thin figures. Girls who were more gender typical, or more similar to girls, were more likely to endorse appearance stereotypes. There was no significant relation among stereotypes about weight change and gender identity. Many girls deemed the fat figure as unattractive, physically restrained, unhealthy, and likely to be bullied and believed that the fat figure should change back to look like a thinner figure. Most girls were aware that exercising, eating healthy, and dietary restriction can lead to weight loss and that food consumption and lack of exercise can lead to weight gain. Findings suggest that certain aspects of weight (anti-fat) and appearance stereotypes are closely linked to gender identity in girls, whereas other aspects of weight stereotypes (e.g., stereotypes about weight change) are so pervasive that they are common in most girls, regardless of their gender identity.

Conceptual knowledge modulates memory recognition of common items: The selective role of item-typicality.

This work examines the influence of stored conceptual knowledge (i.e., schema and item-typicality) on conscious memory processes. Specifically, we tested whether item-typicality selectively modulates recollection and familiarity-based memories as a function of the availability of a categorical schema during encoding. Experiment 1 manipulated both encoding type (categorical vs. perceptual) and item-typicality (typical vs. atypical) in a single Remember-Know paradigm. Experiment 2 replicated and extended the previous study with a complementary source-memory task. In both experiments, we observed that typical items led to more Guess responses, while atypical items led to more Remember responses. These findings support the idea that the activation of a congruent categorical schema selectively enhances familiarity-based memories, likely due to the bypassing of the activated mechanisms for novel information. In contrast, atypical items improved recollective-based memories only, suggesting that their lesser fit with the stored prototype might have triggered those novelty processing mechanisms. Moreover, atypical items enhanced memory in the categorical condition for both item recognition and recollection memories only, suggesting an episodic gain due to inconsistency/novelty. The source memory results gave further credence to the argument that "Remember" judgments were based on truly recollective experiences and presented the same interaction between encoding type and item-typicality observed in recollective-based memories. Overall, the results suggest that the supposedly opposite conceptual knowledge effects actually coexist and interact, albeit selectively, in the modulation of recollection and familiarity processes.

Premise typicality as feature inference decision-making in perceptual categories.

Making property inferences for category instances is important and has been studied in two largely separate areas-categorical induction and perceptual categorization. Categorical induction has a corpus of well-established effects using complex, real-world categories; however, the representational basis of these effects is unclear. In contrast, the perceptual categorization paradigm has fostered the assessment of well-specified representation models due to its controlled stimuli and categories. In categorical induction, evaluations of premise typicality effects, stronger attribute generalization from typical category instances than from atypical, have tried to control the similarity between instances to be distinct from premise-conclusion similarity effects, stronger generalization from greater similarity. However, the extent to which similarity has been controlled is unclear for these complex stimuli. Our research embedded analogues of categorical induction effects in perceptual categories, notably premise typicality and premise conclusion similarity, in an attempt to clarify the category representation underlying feature inference. These experiments controlled similarity between instances using overlap of a small number of constrained features. Participants made inferences for test cases using displayed sets of category instances. The results showed typicality effects, premise-conclusion similarity effects, but no evidence of premise typicality effects (i.e., no preference for generalizing features from typical over atypical category instances when similarity was controlled for), with significant Bayesian support for the null. As typicality effects occurred and occur widely in the perceptual categorization paradigm, why was premise typicality absent? We discuss possible reasons. For attribute inference, is premise typicality distinct from instance similarity? These initial results suggest not.

Penguins, Birds, and Pilot Knowledge: Can an Overlooked Attribute of Human Cognition Explain Our Most Puzzling Aircraft Accidents?

OBJECTIVE: We extend the theory of conceptual categories to flight safety events, to understand variations in pilot event knowledge. BACKGROUND: Experienced, highly trained pilots sometimes fail to recognize events, resulting in procedures not being followed, damaging safety. Recognition is supported by typical, representative members of a concept. Variations in typicality ("gradients") could explain variations in pilot knowledge, and hence recognition. The role of simulations and everyday flight operations in the acquisition of useful, flexible concepts is poorly understood. We illustrate uses of the theory in understanding the industry-wide problem of nontypical events. METHOD: One hundred and eighteen airline pilots responded to scenario descriptions, rating them for typicality and indicating the source of their knowledge about each scenario. RESULTS: Significant variations in typicality in flight safety event concepts were found, along with key gradients that may influence pilot behavior. Some concepts were linked to knowledge gained in simulator encounters, while others were linked to real flight experience. CONCLUSION: Explicit training of safety event concepts may be an important adjunct to what pilots may variably glean from simulator or operational flying experiences, and may result in more flexible recognition and improved response. APPLICATION: Regulators, manufacturers, and training providers can apply these principles to develop new approaches to pilot training that better prepare pilots for event diversity.

Evidence for Shifting Cognitive Strategies when Icons Appear in Unexpected Locations.

OBJECTIVE: The present study examines the cognitive effects of placing icons in unexpected spatial locations within websites. BACKGROUND: Prior research has revealed evidence for cognitive conflict when web icons occur in unexpected locations (e.g., cart, top left), generally consistent with a dynamical systems models. Here, we compare the relative strength of evidence for both dual and dynamical systems models. METHODS: Participants clicked on icons located in either expected (e.g., cart, top right) or unexpected (e.g., cart, top left) locations while mouse trajectories were continuously recorded. Trajectories were classified according to prototypes associated with each cognitive model. The dynamical systems model predicts curved trajectories, while the dual-systems model predicts straight and change of mind trajectories. RESULTS: Trajectory classification revealed that curved trajectories increased (+11%), while straight and change of mind trajectories decreased (-12%) when target icons occurred in unexpected locations (p < .001). CONCLUSION: Rather than employing a single cognitive strategy, users shift from a primarily dual-systems to dynamical systems strategy when icons occur in unexpected locations. APPLICATION: Potential applications of this work include the assessment of cognitive impacts such as mental workload and cognitive conflict during real-time interaction with websites and other screen-based interfaces, personalization and adaptive interfaces based on an individual's cognitive strategy, and data-driven A/B testing of alternative interface designs.

The typicality effect in basic needs.

According to the so-called Classical Theory, concepts are mentally represented by individually necessary and jointly sufficient application conditions. One of the principal empirical objections against this view stems from evidence that people judge some instances of a concept to be more typical than others. In this paper we present and discuss four empirical studies that investigate the extent to which this 'typicality effect' holds for the concept of basic needs. Through multiple operationalizations of typicality, our studies yielded evidence for a strong effect of this kind: (1) Participants tended to recall the same core examples of the concept in a free-listing task. (2) They judged some basic needs to be more typical than others. (3) The items that were judged to be more typical were listed more frequently in the free-listing task. (4) These items were listed earlier on in the free-listing task. (5) Typical basic needs, as well as non needs, were classified faster than atypical basic needs in a reaction time study. These findings suggest that the concept of basic needs may have a non-classical (e.g., exemplar or prototype) structure. If so, the quest for a simple and robust intensional analysis of the concept may be futile. Supplementary Information: The online version contains supplementary material available at 10.1007/s11229-022-03859-9.

Sexual Dimorphism in Facial Contrast: A Case from Central Africa.

Apart from morphological differences, male and female faces also vary in color, especially in overall lightness and facial contrast, i.e., the contrast between the luminance and color of facial features (eyes, lips, or brows) and luminance and color of the surrounding skin. In many populations, it has been demonstrated that women tend to be lighter than men. Other differences were found in facial contrast: women have a higher contrast between the lightness of their eyes and lips and the surrounding skin. Manipulation of this contrast in an artificial genderless face can result in a masculine or feminine appearance. So far, however, this phenomenon has been studied mostly in Euro-American and East Asian samples, with little evidence from populations with darker facial tone. We explored natural sexual dimorphism in both facial contrast and lightness in an African, namely Cameroonian, sample, and compared it with results for a European, in particular Czech, population. Our findings showed that sexual differences in luminance contrast of eyes and brows were in both studied populations similar but in the Cameroonian sample, significant difference in lips contrast was absent. These results indicate that sex differences in facial contrast are a side effect of the sex differences in skin color and can be used as a proxy for skin color perception.

Using concept typicality to explore semantic representation and control in healthy ageing.

Successful use of conceptual knowledge entails the assembling of semantic representations and control processes to access the subsets of knowledge relevant in each situation. Research has suggested that representation and control mechanisms interact to support categorization. Notably, depleted representations in semantic dementia and disrupted control processes in semantic aphasia impair categorization of atypical concepts. Yet, it remains unclear how knowledge accumulation and control decay in healthy ageing impact categorization. To address this question, we compared young and older adults' performance in a categorization task of items varying in concept typicality. Critically, older adults were more accurate in categorizing atypical concepts than the younger counterparts, as indicated by the interaction between group and typicality. Moreover, the elderly outperformed the younger in categorizing atypical concepts that were also less familiar. Thus, the decay in semantic control observed along ageing did not significantly hinder the categorization of atypical items. Our data suggest that, relative to young adults, older adults possess enriched conceptual knowledge, which supports retrieval of the category-related features needed for categorizing atypical and less familiar exemplars.

Popularity and Gender Prototypicality: An Experimental Approach.

Despite the growing scientific understanding of peer popularity, there are few theories that explicitly address the development of peer popularity in adolescence. The studies reported here present a preliminary test of the theory that popularity is associated with gender prototypicality. Popularity is associated with physical attractiveness, as well as with attributes (e.g., athletic involvement for boys, having stylish clothes for girls) that often reflect gender-based expectations. After being exposed to either a high school popularity priming condition or a neutral control condition, 1st-year college students rated photographs (Study 1, N = 368, 34% male, 66% female; Mage = 19.30, SD = 1.78, range 17-35), vignettes (Study 2, N = 249, 16.4% males, 83.2% females, 0.4% other; Mage = 18.71, SD = 2.31, range 17-40), and social media profiles (Study 3, N = 218, 30.3% male, 69.3% female, 0.5% other; Mage = 19.40, SD = 2.31, range 18-39) depicting gender-typical and gender-atypical adolescents' appearance and interests on a number of popularity-related characteristics. These results indicated that gender prototypicality in both appearance and interests is associated with popularity.

Encouraging Gender Conformity or Sanctioning Nonconformity? Felt Pressure from Parents, Peers, and the Self.

Expectations regarding gendered behaviors are understood to emanate from many sources, including parents, peers, and the self but there has been little research directly comparing these three sources of pressure. The present study assessed felt pressure regarding masculine and feminine stereotypic behaviors and compared pressure from parents, peers, and the self and how these sources are associated with self-perceived gender typicality. Participants (N = 275; 53.09% female; grade 7 M = 12.35 years; grade 9 M = 14.3 years; 71% Anglo-Celtic) were recruited from independent schools in Sydney. Felt pressure from the self to engage in masculine behavior was greater than felt pressure from parents or peers. Male adolescents reported higher felt pressure to conform to gender conforming behavior from parents and the self and pressure to avoid gender nonconforming behavior was greatest from peers. Female adolescents reported felt pressure to conform to gender nonconforming behaviors and this pressure was the strongest from the self. Additionally, for both genders, felt pressure from the self was most consistently associated with gender typicality. These findings highlight the importance of self-expectations for gender cognitions relating to both masculine and feminine behaviors.