Research has shown that White adults display a race-based size bias (Freiburger et al., 2023; Wilson et al., 2017); they tend to perceive Black men as stronger, taller, heavier, and more capable of harm than White men of the same physical size (Wilson et al., 2017). This bias has serious implications. Black men are subject to more frequent stops by the police when they are perceived as tall (Hester and Gray, 2018). Also, non-Black participants' perceptions of Black men's physical size predict their perceptions of harm and use of police force justifications (Wilson et al., 2017). White adults also represent negative interactions as more violent when performed by a Black person (Duncan, 1976), perceive non-dangerous objects (e.g., phone) as more threatening when they are paired with Black men, and associate weapons more quickly with Black men than with White men (Correll et al., 2002; Payne, 2001; Todd et al., 2021). The current study is an initial investigation into the presence of a race-based size bias in children, adolescents, and young adults.

Research focused on a race-based size bias in adults is situated in research on adult social cognition investigating the causes and consequences of representing Black men as threatening and more capable of harm (Correll et al., 2002; Hester and Gray, 2018; Holbrook et al., 2016; Payne, 2001; Todd et al., 2021; Wilson et al., 2017). Although there are several distinct mechanisms that contribute to racially-biased beliefs, Wilson et al. (2017) investigated the possibility that race-based threat perceptions derive from differences in how people perceive the physical size and formidability of Black men. Across a series of studies, the researchers made non-Black participants (though predominantly White) provide a series of size and threat-related judgments for Black and White targets. Across these studies Wilson and colleagues found consistent evidence for a race-based size bias. These biased responses were evident when bodies were presented with real photographs of Black and White faces – taken from individuals who were matched in actual physical size - as well as images of bodies that masked the race of the target but were presented alongside distinctively Black or White names. While many of the studies conducted by Wilson et al. (2017) only employed non-Black perceivers, one study specifically assessed a potential race-based size bias in Black participants. They found that Black participants also judge Black men are physically stronger, but not more threatening, than White men of the same size showing that a race-based strength bias extends beyond White perceivers. Moreover, this bias appears to be specific to judgments about Black men, as Johnson and Wilson (2019) found that White, Asian, and Black perceivers rated Black men as stronger than both White and Asian men.

Recent evidence also suggests that a race-based size bias is not limited to White adults' judgments about Black adult males. Freiburger et al. (2023) showed that White adults also judge Black adolescents as taller than White adolescents when judging real photographs of adolescents that were matched in height as well as computer-generated faces that were matched in all aspects except racial phenotypicality. White adults also rated Black adolescents in real photographs as less innocent than White adolescents. Finally, manipulating the anger expressions of faces – a visible signal of threat – influenced White adults height ratings only for White adolescents; while participants exhibited a baseline bias to rate Black adolescents as taller than White adolescents, adding a threat signal via an angry facial expression increased the size ratings for White adolescents only. These findings suggest that White adults have a bias to perceive Black individuals, as young as adolescents, as both taller and more threatening than White individuals.

Taken together, this area of research suggests that rather than being rooted in broadly-biased perceptions of out-group individuals, the race-based size bias appears to be rooted in a conceptual bias in how individuals specifically represent Black men. Although most research has been conducted with White participants, some evidence suggests that the bias to represent Black adult and adolescent males as physically larger (though not threatening) is also present in non-White adults. In the current study, we investigate the presence of the race-based size bias in children, adolescents, and young adults. To our knowledge, there has been no research on how this bias across development. Do children similarly display a bias to perceive Black people as taller, heavier, and stronger than White people, or is this bias specific to adults' racial biases? Does this bias present itself differently across development? Given prior research provides some evidence that this bias is present in individuals from a range of racial backgrounds, we similarly investigate these questions in a sample of White and non-White participants; we also return to questions concerning the nature of a race-based size bias across White individuals and non-White individuals in the General Discussion.

Characterizing the presence of a race-based size bias across development is important for several reasons. First, developmental research can contribute to our theoretical understanding of the bias in adults. For example, do children initially represent Black individuals as threatening, and then map these race-threat biases to representations of size? Or, do children initially represent larger individuals as threatening, and then apply this size-threat bias to Black individuals as they acquire other stereotypes across development? Knowledge that one aspect of the bias emerges first developmentally may provide insight into how this bias operates cognitively in adults. Second, knowledge of when the bias emerges may help us identify the potential age-appropriate intrinsic (e.g., cognitive mechanisms) and extrinsic (e.g., environmental inputs) factors that we may consider in explaining how the bias develops. Finally, developmental research can inform interventions to potentially mitigate this bias. The absence of developmental research precludes us from addressing any of these questions. We begin to discuss how the findings of the current study can begin to speak to these questions in the General Discussion.

Although no prior developmental research has directly investigated the association between representations of race and physical size, researchers have investigated whether components of a race-based size bias are present across development. We briefly review below findings on children's reasoning related to race and physical size.

A large body of literature has shown that children (predominantly White children in many studies) are attuned to race – cued by differences in skin color – throughout childhood, and this early social cognition is influenced by the racial and ethnic diversity of children's environments (Anzures et al., 2013; Dunham et al., 2008; Mandalaywala and Legaspi, 2023; Sherman et al., 2013). Children who live in less racially diverse environments hold some essentialist beliefs about race and exhibit explicit in-group favoritism based on race (Aboud, 2003; Gedeon et al., 2021; Hirschfeld, 2008; Mandalaywala et al., 2019; McGlothlin and Killen, 2006, 2010; Pauker et al., 2016a,b; Rutland et al., 2005); children also rate both ambiguous and unambiguous actions as meaner and more threatening when they are performed by a Black actor than a White actor (Lawrence, 1991; McGlothlin and Killen, 2010; Sagar and Schofield, 1980). Implicit pro-White and anti-Black attitudes are present as young as 6 years of age (Baron and Banaji, 2006), and these attitudes lead children to apply stereotypic information throughout childhood (Bigler and Liben, 2007). Finally, children are also aware of racial stereotypes by at least age 6, and this awareness increases throughout elementary school (Ambady et al., 2001; Castelli et al., 2009; Devine, 1989; McKown and Strambler, 2009; McKown and Weinstein, 2003; Sierksma et al., 2022); explicit endorsement of negative stereotypes is more likely to emerge in adolescence (Copping et al., 2013; Evans et al., 2011; Okeke et al., 2009; Rowley et al., 2007). Taken together, this research suggests that children display early social biases based on race, but that endorsement of specific racial stereotypes emerges later toward adolescence (Devine, 1989; Sierksma et al., 2022).

With respect to children's representations of physical size, research has shown that children recognize body size and its relation to dominance early in infancy and throughout childhood (Lourenco et al., 2016; Mascaro and Csibra, 2012; Terrizzi et al., 2019; Thomsen et al., 2011); for example, children have an expectation that individuals with larger body sizes should prevail in a dominance interactions. Children also form social preferences based on body size (i.e., preferring people with smaller bodies) and reason about the causes of height and weight by age four (Carvalho et al., 2021; Cramer and Steinwert, 1998; Harrison et al., 2016; Musher-Eizenman et al., 2004; Penny and Haddock, 2007; Raman, 2014, 2017). Yet, no prior research, to our knowledge, has investigated whether children form associations between physical body size and race. Taken together, developmental research demonstrates that children represent the individual characteristics associated with a race-based size bias by at least the preschool years. Yet, when these representations are linked in a manner similar to a race-based size bias seen in adulthood is unknown.

The current study investigated the presence of a race-based size bias in children, adolecents, and young adults. Child (6- to 11-years), adolescents (12- to 17 years), and adult (college students, over 18 years of age) participants were introduced to male characters, who varied in race (Black, White) across trials and engaged in negative interactions with an unseen social partner. We then assessed how participants' age and the character's race influenced judgments of the character's physical size (i.e., strength, height, weight).

We faced several methodological decisions when considering how to investigate the race-based size bias across several age ranges. Crucially, the current study aimed to employ the same stimuli and measures across different age groups to begin to tackle this challenge, which led to a series of careful, methodological choices.

First, we began by deciding the age of the targets. Previous research focusing on the race-based size bias have used both adolescent and adult faces, but also only had adult participants (Freiburger et al., 2023; Johnson and Wilson, 2019; Wilson et al., 2017). Because we recruited participants across many age ranges - children, adolescents, and adults – we believed using child stimuli would be easier for children to gauge in terms of size estimation. Previous research has shown that children (6–14 years of age) are relatively accurate at estimating their own body size (Gardner et al., 1999), but there is no research to our knowledge about children estimating adult body sizes. Intuitively, we hypothesized that children would use only the higher end of a scale estimating all adults to be large and tall since adults are much larger than our child participants. Additionally, adults are easily able to make judgments about Black and White children, and may generalize stereotypes about Black adults to young Black children (Todd et al., 2016). Thus, using stimuli of children for a study with child, adolescent, and adult participants seemed appropriate.

Second, we decided to use cartoon faces, rather than photographs of real faces, to depict the targets in our study to ensure we had control over the visual features present in the faces. While some researchers have criticized the use of cartoon and computer-generated faces in social cognitive development studies, and questioned their generalizability (Gaither et al., 2019; Packer and Moreno-Dulcey, 2022; Rakoczy, 2022; Wellman and Yu, 2022), other researchers argue that children often represent non-human stimuli as human and that this methodological choice offers more experimental control and makes it easier for researchers to study and understand children (Revencu and Csibra, 2024). This allowed us to control for the size of the faces while still ensuring participants would categorize the stimuli as “Black” and “White” categorically.

Our final methodological decision concerned how to assess the strength, height, and weight biases psychometrically. We assumed that children would have difficulty providing numerical judgments of heights and weights. Therefore, we opted to use Likert scales that could be used across all age ranges in this study.

The predictions for our study were primarily exploratory given that no research has been conducted on the race-based size bias developmentally. Given that children represent physical size, exhibit racial preferences, and display out-group biases as young as the preschool years, a race-based size bias may be present even in the youngest children tested in this study. However, given that children's stereotype awareness and endorsement increase across childhood, a race-based size bias may also emerge slowly across development, with little evidence of a race-based size bias at younger ages and increasing evidence in later childhood and adolescence. The findings from the current study can help begin to distinguish these alternative hypotheses.

The current study aimed to investigate the presence of a race-based size bias across childhood to early adulthood. We presented children, adolescents, and college students with Black and White male characters engaging in negative events and had them rate the character's height, weight, and strength. Although we did not find consistent evidence for ratings of height and weight, we did find evidence for age-related differences in ratings of strength. We found that (1) event type, but not race, influenced children's strength ratings, (2) race and event type independently influenced adolescents' strength ratings, and (3) the influence of race on adults' strength ratings varied by event. The findings suggest that a racial bias related to physical strength operates differently across development. Although children represent both race and physical size, these representations are not associated with each other similar to the race-based size biases seen in adults. Rather, children may initially form associations between only physical strength and only specific kinds of events, incorporating representations of race into these associations later in development as they begin to enter adolescence.

Although the current study did not investigate the specific factors that explain how this bias may emerge over development, characterizing its presence at different ages is an important first step and motivates more targeted research on the developmental changes identified in the current study. For instance, social-cognitive developmental research suggests that children's representations of social status play a particularly important role in children's developing representations of social groups (Baron et al., 2016; Dunham et al., 2008; Mandalaywala et al., 2020; Rhodes and Baron, 2019; Rhodes and Mandalaywala, 2017). Yet, the findings from the current study suggest that a race-based strength bias might not follow immediately from social status representations; although research has shown that children can reason about social status at the youngest ages tested in the current study, a race-based strength bias did not emerge in the current study until adolescence.

One alternative possibility is that a race-based strength bias is linked to awareness of the specific stereotype that Black men are large. Given research showing that endorsement and knowledge of stereotypes increase with age (Copping et al., 2013; Evans et al., 2011; Okeke et al., 2009; Rowley et al., 2007; Sherman et al., 2013; Sierksma et al., 2022), the current findings might similarly support conclusions that individuals do not become aware of the stereotypes needed to support a race-based strength bias until adolescence. Another possibility, not mutually exclusive from the first, is that exposure to media that present stereotyped characters might be tied to the onset of a race-based strength bias (Ellithorpe and Bleakley, 2016). Finally, it is possible that the ability to represent both environmental context and character race equally – as well as to integrate those representations – improves with age, leading to the age-related onset of the race-based strength bias seen in the current study. The current findings suggest that investigating how a race-based strength bias may be related to changes in other social cognitive mechanisms or changes in environmental contexts in childhood and early adolescence is an important area for future research.

We also found that event context significantly influenced participants' rating across development: characters who engaged in physical events were perceived as stronger than those in social events. This may seem intuitive, as engaging in physical actions (e.g., pushing someone) may require physical strength, whereas socially excluding others does not. Yet, the adult strength results also show an important interaction where Black characters who engaged in social events were rated as stronger than White characters who engaged in social events, but the same was not true for physical events. Although not statistically significant, the adolescent data showed a similar descriptive pattern. We believe this may be due to the potential ambiguity provided by a social event. In a physical event in which someone is pushed to the ground, the actor is clearly strong and mean. However, in a social event, participants may be more likely to think about other potential situational factors (e.g., saving the seat for someone else), and provide a situational explanation for White characters more than for Black characters; conversely, participants may be more likely to provide a personal attribution for Black characters over White characters. In short, the ambiguity present in social events may have provided conceptual space for racial biases to emerge. We also note that while we aimed to only investigate participant's perceptions of negative events, about 5% of the trials were perceived positively by participants. Although this subset of data was too small to assess in the current study, it does raise important questions about how the valance of an event, as well as the degree to which participants rate such valence, influence their representations of physical strength. Given we did not make any a priori predictions related to valence or event type, these are merely speculative interpretations of the data and motivate further studies to explore these questions.

Although we found evidence for the presence of a race-based strength bias starting in adolescence, we did not find a similar bias at any age in height or weight judgments. We believe that there may be both methodological and conceptual explanations for these differences. Although some prior studies on adult participants have found evidence of a race-based size bias using height and weight judgments (e.g., Wilson et al., 2017), there is at least one instance of a race-based size bias emerging only for strength judgments (Johnson and Wilson, 2019). The current findings similarly suggest that a race-based size bias may be associated more closely with physical strength than height and weight. Given the inconsistent nature of the height and weight ratings for this bias, we believe that the strength bias may be the most robust amongst the different size variables.

It is also possible that the use of Likert scales may have masked the presence of a race-based strength bias in children. Other prior research has explored adults' perceptions of real adult and adolescent faces, rather than cartoon child faces, and has used slider scales or estimates of height and weight (Freiburger et al., 2023; Johnson and Wilson, 2019). We based our measurement choice on (1) knowledge that prior research with adults has used both self-reported or continuous scales and (2) the assumption that the emergence of a race-based size bias is conceptually tied to developmental changes in children's knowledge about physical size. Although how one defines strength (i.e., what counts as stronger vs. weaker), for example, likely changes over development, we assumed that children would still be capable of reporting general differences in strength. Similarly, children may not know how specific height and weight metrics make someone tall or heavy, we assumed they could report relative differences. Nonetheless, future research may employ different kinds of measurements to characterize how they relate to each other as well as change across development. It is possible that children may be more likely to display a race-based strength bias when making forced choice decisions (e.g., who is stronger/taller/heavier?).

We also acknowledge that participants made size judgments for child characters rather than adult characters. Given we were interested in the presence of this bias across development, we opted to use child cartoon characters in this study to increase the likelihood that children had sufficient knowledge to inform their ratings of physical size and to allow us to control the physical size of the characters across trials. However, as noted above, representations of physical size certainly differ based both on the age of the character as well as the age of the participants. Future studies may wish to assess whether race-based size biases emerge differently based on the age of the characters for these judgments. The developmental trajectories for a race-based size bias for child characters may differ from a bias for adolescent or adult characters.

Additionally, we understand that the difference in testing modality across age groups may be a confound in the current study. While children were tested on-line and in person, adolescents were tested only in person, and adults were tested only online. Given that our follow-up analyses did not reveal any significant differences between testing modalities in our child participants – the only group that had been tested with multiple modalities – we do not believe that testing modality can explain the age-related differences found in the current study. Instead, we believe the range of developmental factors discussed above are likely to better explain the current findings. Yet, future studies may still aim to keep testing modality constant across age groups.

Although our overall sample was racially diverse, it is important to acknowledge that we did not recruit enough participants across racial groups to assess for how participant's race may influence their representations of height, weight, and strength. Understanding whether and how these biases differ across participants' racial backgrounds is an important research question. For instance, as noted in the Introduction, Wilson et al. (2017) showed that while both Black and White adults exhibited a race-based strength bias, Black adults did not similarly represent Black characters as more threatening. Although we have used generic language to describe our findings, we acknowledge that a complete explanation of how a race-based strength bias emerges in individuals across development will also need to consider whether, when, and how such biases may differ across individuals from different racial backgrounds.

Finally, while the current findings suggest that there may be important social-cognitive or environmental factors that contribute to change in a race-based strength bias between childhood and adulthood, we are unable to make specific recommendations for educational or psychological intervention. Instead, this study should be considered an initial investigation of how a race-based strength bias begins to emerge across development. Following additional research investigating how social, cognitive, and environmental factors influence this bias at different ages, as discussed above, future research may then focus on designing and implementing psychological or educational interventions that may mitigate this bias across development.

In summary, this study represents an initial investigation into the presence of a race-based size bias across childhood to young adulthood. The findings suggest that although children represent race and physical size early in development, and form negative attitudes about out-groups early in development, these representations do not automatically result in a race-based size bias seen in adulthood. Rather, our findings suggest that children's representations of physical size may initially be focused on representations of strength and tied more closely to specific kinds of events, only later becoming similar to the racial biases seen in adults. Future studies characterizing the development of this bias, as well as the factors that explain its potential emergence, will be important areas of future research.