Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Sep 1.
Published in final edited form as: Body Image. 2019 Jun 22;30:107–113. doi: 10.1016/j.bodyim.2019.06.005

You can buy a child a curvy Barbie doll, but you can’t make her like it: Young girls’ beliefs about Barbie dolls with diverse shapes and sizes

Jennifer A Harriger a,*, Lauren M Schaefer b, J Kevin Thompson c, Li Cao d
PMCID: PMC6857835  NIHMSID: NIHMS1055799  PMID: 31238275

Abstract

This study utilized Mattel’s new line of Fashionista Barbie dolls to examine attitudes about body shape and size in a sample of young girls. A total of 84 girls, 3–10 years of age, were asked to assign positive or negative traits to Barbie dolls which varied in size and shape (original, tall, petite, and curvy). Participants also answered questions about their preferences for the dolls and completed measures of body dissatisfaction. Results generally demonstrated greater negative attitudes towards the curvy Barbie doll and more positive attitudes towards dolls with a thinner body size/shape (i.e., original, tall, and petite dolls). Girls identified the curvy Barbie as the doll they least wanted to play with. Additionally, girls with higher levels of body dissatisfaction demonstrated less negative attitudes towards the original doll. Overall, findings demonstrate a preference for thin bodies and aversion towards larger bodies among young girls. Further, findings suggest that the simple availability of body-diverse dolls may not be a powerful enough intervention to overcome harmful weight attitudes, and highlight the importance of continued efforts to encourage exposure to and acceptance of diverse body shapes and sizes in young children.

Keywords: Barbie, Children, Body dissatisfaction, Weight bias, Thin ideal

1. Introduction

Positive attitudes towards thin bodies (e.g., internalization of the thin ideal, or endorsement of thinness as one’s preferred body type) and negative attitudes towards larger bodies (e.g., weight bias, or negative attitudes and discrimination toward larger individuals based on their body weight) are common among adult women, and are associated with deleterious mental and physical health outcomes including body image disturbance, disordered eating, low self-esteem, stress, anxiety, and depression (Dittmar & Howard, 2004; Pearl, White, & Grilo, 2014; Pearl, White, & Grilo, 2014; Puhl & Heuer, 2009; Thompson & Stice, 2001). Given this, research has attempted to better understand the development of these experiences in order to identify appropriate prevention and intervention approaches. A large body of work suggests that appearance and weight attitudes are culturally informed (Swami et al., 2010), developing in part in response to dominant sociocultural messages venerating certain body types over others (Veldhuis, tel Poel, Pepping, Konijn, & Spekman, 2017). Consequently, a growing body of work has sought to examine the developmental timeline for body size and weight attitudes, as well as the specific sources and impact of sociocultural weight-related messages (e.g., Cramer & Steinwert, 1998; Kornilaki, 2015; Latner, Simmonds, Rosewall, & Stunkard, 2007)

Research examining potentially harmful attitudes about body size and weight among youth may provide valuable information regarding the ages at which these beliefs begin to develop. Importantly, investigators are frequently unable to use standard self-report measures of thin internalization and weight bias, which have been developed and validated for use with adults, to assess weight attitudes among youth. Instead, these teams have more commonly employed a variety of developmentally-adjusted tasks to assess children’s attitudes about body size and weight (Harriger, Calogero, Witherington, & Smith, 2010; Musher-Eizenmann et al., 2004). For example, research with preschool children typically employs methods which require no reading skills and little to no verbal responses from participants. In one method that has been used with children as young as 2.5 years of age, participants are shown two to three line-drawn figures of differing sizes (e.g., thin, average, or large) and read a list of words (e.g., happy, sad, has friends, has no friends). Participants are then asked to point to the figure that looks like the word that is being read (Harriger, 2015; Turnbull, Heaslip, & McLeod, 2000; Wright & Bradbard, 1980). Findings from this work demonstrate that children as young as three years of age reliably display a preference for thinness (i.e., assign more positive characteristics to thinner figures) and negative attitudes towards higher weights (i.e., assign more negative characteristics to larger figures), suggesting that attitudes about body size and weight develop very early in life (e.g., Harriger, Trammell, Wick, & Leudke, 2019; Spiel, Paxton, & Yager, 2012).

Although the above findings have been widely cited, some researchers have speculated that the utilized tasks are not developmentally appropriate for young children. For example, Meers and colleagues (2011) suggested that line drawings may be too abstract for very young children and recommended utilizing less abstract stimuli in future work with youngsters. Echoing these concerns, Worobey and Worobey (2014) argued that the use of concrete objects (i.e., dolls) may minimize the potential for methodological bias. However, to date, very few studies have utilized this approach, which limits confidence in earlier findings using more abstract stimuli.

In addition, the developmental trajectory of weight attitudes is not well understood. In samples of children ages 3–5 (Harriger et al., 2019), 5–10 (Kornilaki, 2015) and 9–17 (Klaczynski, Daniel, & Keller, 2009), weight bias has been shown to increase with age. However, other studies among 4- to 6-year-olds (Holub, 2008) and 10- to 13-year-olds (Latner et al., 2007) report no significant age differences in weight attitudes. While less is known about the developmental trajectory of beliefs about body size and shape, researchers have noted that children who endorse greater negative attitudes towards people at higher weights (i.e., weight bias) also experience higher levels of depression, anxiety, low self-esteem, body dissatisfaction, and disordered eating (Harriger & Thompson, 2012; Paxton & Damiano, 2017; Puhl & King, 2013). Additionally, youth who internalize weight bias beliefs (i.e., experience self-directed weight stigma) are more likely to report these same consequences regardless of weight status (Pearl & Puhl, 2014; Zuba & Warschburger, 2018). Therefore, additional research regarding the developmental timing of weight attitudes is warranted, as such work may help to identify critical periods during which these harmful beliefs develop and may be most promisingly intervened upon.

Further, examination of factors that may contribute to the development of attitudes about body size and shape in youth is a worthwhile area of study as such work could help to identify meaningful targets for interventions. Although the development of attitudes about weight and body size is likely multifactorial (Suisman et al., 2012), a large body of evidence suggests that exposure to images of thin, highly idealized female bodies contributes to body dissatisfaction and increased thin-ideal internalization among adult women (Frederick, Daniels, Bates, & Tylka, 2017; Groesz, Levine, & Murnen, 2002; Levine & Harrison, 2004). In a parallel process, one way in which young girls may begin to learn about and internalize societal weight ideals is through frequent exposure to very thin idealized dolls common in their play (Boyd & Murnen, 2017; Dittmar, Halliwell, & Ive, 2006; Kuther & McDonald, 2004). Within a crowded and competitive marketplace, the Barbie doll holds an iconic status as the best-selling fashion doll world-wide. Mattel estimates that 90% of girls between the ages of 3 and 10 own at least one Barbie doll (Dittmar et al., 2006; Kuther & McDonald, 2004). It is further estimated that over one billion Barbie dolls have been sold in over 150 countries, with three Barbie dolls being sold every second (In Depth: Barbie by the Numbers, 2009; Vintage Barbie Struts Her Stuff, 2006). Mattel markets the doll as both an inspirational and aspirational figure, encouraging young girls to imagine themselves taking on the roles and attributes of its popular doll (Pedersen & Markee, 1991; Turkel, 1998). Indeed, for many youngsters Barbie represents the pinnacle of femininity and attractiveness (Kuther & McDonald, 2004).

A key aspect of Barbie’s appearance relates to her slender physique. It is estimated that if Barbie were life-size she would be 5′9″ and a Size 2 in the United States (Dittmar et al., 2006). Further, it is suggested that Barbie would lack adequate adipose tissue to support menstruation (Winterman, 2009). Given Barbie’s unrealistically thin and seemingly unhealthy figure (Lind & Brzuzy, 2008), a number of researchers have suggested that Barbie dolls may promote harmful weight attitudes including thin-ideal internalization in young girls (Tiggemann, 2011, Dittmar, 2012). Experimental exposure studies seem to provide partial support for these concerns. For example, Dittmar et al. (2006) presented girls ages 5–8 with a picture book containing an image of a Barbie or a doll representing a larger and more realistic body size, the Emme doll, whose proportions approximate a Size 16 in the United States. They found that exposure to the images of Barbie resulted in higher levels of body dissatisfaction in their younger participants (i.e., girls age 5.5–7.5), but not their older participants (i.e., girls age 7.5–8.5). Notably, however, the study did not use real dolls, and therefore the images may have been less realistic than actual dolls. As a follow-up, Anschutz and Engels (2010) examined differences in body image and eating behaviors among 6- to 10-year-old Dutch girls exposed to either a thin doll (i.e., Barbie doll or Tyler doll) or an average sized doll (i.e., Emme doll). Although they found no effect of doll exposure on body satisfaction, they did observe an effect on subsequent ad libitum food intake, with girls who were exposed to the thin dolls consuming significantly less than girls exposed to the average sized doll. This effect was not moderated by age. Finally, Rice, Prichard, Tiggemann, and Slater (2016) found that experimental exposure to Barbie promoted higher levels of thin-ideal internalization, but not body esteem or body dissatisfaction, compared with exposure to My Little Pony in a sample of 5- to 8-year-old girls. Again, age did not moderate this effect.

Overall, these results suggest that exposure to unrealistically thin dolls may negatively influence children’s feelings about their own bodies and their eating behaviors, and that age may moderate the impact of doll exposure. In response, some researchers and advocates have encouraged toy manufacturers to mass produce dolls with more diverse body types. These voices suggest that increasing children’s exposure to body-diverse dolls may help to lessen the potentially harmful effects of repeated exposure to a thin body ideal and allow an opportunity for children to develop greater flexibility in body attitudes (Dittmar et al., 2006).

In February 2016, Mattel released a new line of Barbie Dolls called the Fashionista line. The new Barbies come in a variety of body types (i.e., curvy, tall, and petite). Original Barbie is thin, with an hourglass figure and highly pronounced “thigh gap” (i.e., space between the dolls thighs), signaling very low adiposity. Tall Barbie is similarly thin, but with longer legs and torso, and a less pronounced hourglass shape. Relative to the original doll, petite Barbie has shorter legs and torso. She retains a thin figure and hourglass shape, but has a less pronounced “thigh gap.” Curvy Barbie is similar in height to petite Barbie, but has a more rounded stomach and thicker legs with no gap between her thighs, signaling greater adiposity. Thus, while the dolls differ considerably in size and shape, the original, tall, and petite dolls share similar levels of thinness, while the curvy doll portrays a fuller figure.

Although the introduction of varied body types into the Barbie doll lineup represents a clear response to calls for greater body diversity in children’s toys, it is notable that both the tall and petite dolls remain extremely thin. While the curvy doll appears to possess a more rounded figure, she remains considerably slimmer than the average American woman. More specifically, it is estimated that tall Barbie would be 5′11″ with a Size 4 waist in the United Kingdom and Size 2 hips in the United States; petite Barbie would be 4′11″ with a Size 2 waist and Size 0 hips (US size 0/2); curvy Barbie would be 5′6″ with a Size 6/8 waist and Size 8 hips (US size 4/6; Bates, 2016). In contrast, the average American woman is 5′4″ and a Size 16 (Christel & Dunn, 2016).

While the new line of Barbie dolls presents children and families with a wider array of options in terms of doll size and shape, no study to date has examined children’s responses to the dolls. Such work may provide valuable information related to the acceptability and potential impact of the new Fashionista doll line. Further, the release of this new line of Barbie dolls presents researchers with a valuable opportunity to assess children’s beliefs about body size and shape using concrete dolls that they have an opportunity to interact with on a daily basis.

Building on previous work, the primary purpose of the current study was to examine young children’s attitudes about body shape and size using the original and new Fashionista Barbie dolls. As extant research using a relatively narrow age range of participants suggests a potential moderating impact of age, the current study sought to include participants from a broader age range (i.e., 3–10 years old) reflecting the target audience for Barbie sales. In addition, we examined the associations between children’s age, perceptions of their own body size, body dissatisfaction, and attitudes about body shape and size. It was hypothesized that participants would attribute more negative characteristics to the curvy Barbie and more positive characteristics towards the original Barbie doll, and that this pattern of responding would increase with age. It was also hypothesized that body dissatisfaction would be related to more positive attitudes towards the original doll and more negative attitudes towards the curvy doll. Although the tall and petite Barbie dolls varied in “height,” the estimated dress size for these dolls was similar to that of the original Barbie (i.e., Size 2). As these dolls appeared to vary along a dimension of height, while maintaining a thin figure, specific hypothesizes were not forwarded and exploratory analyses were conducted to examine children’s beliefs about these dolls.

2. Method

2.1. Participants

Participants included 84 girls, ages 3–10 (M = 6.45, SD = 2.43), attending preschools and elementary schools (kindergarten to 5th grade) in southern California. Fifty-eight percent of participants were Caucasian, 9.5% were Asian American, 9.5% were Hispanic, 7.1% were African American, 1.2% were Native American, and 14.3% were other. The majority of the participants (76.2%) reported owning at least one Barbie doll (M = 8 Barbie dolls).

2.2. Materials

2.2.1. Stimuli

The four dolls (original, petite, tall, and curvy) had identical facial features including skin, eye, and hair color but differed in body size and shape (see details in Introduction). To ensure that the stimuli only differed in body shape/size, four additional identical dolls were purchased, and their heads were switched with the heads of the stimulus dolls. The Barbie dolls were dressed in identical bikinis to expose their body shapes (see Fig. 1).

Fig. 1.

Fig. 1.

Open in a new tab

Original, petite, tall, and curvy Barbie dolls.

2.2.2. Adjective attribution task

Participants were presented all four Barbie dolls arranged in a random order. The researcher said to the child “Point to the Barbie that you think is -----.” In total, the children were asked to do this for 10 different adjectives (five positive adjectives and five negative adjectives). The positive adjectives were: happy, smart, has friends, pretty, helps others. The negative adjectives were: sad, not smart, has no friends, not pretty, mean. This list of words was compiled from previously published work utilizing similar techniques (Cramer & Steinwert, 1998, Harrigeret al., 2010; Musher-Eizenman, et al., 2004; Worobey & Worobey, 2014).

2.2.3. Barbie preference

While continuing to view the four Barbie dolls, participants were asked to respond to two questions: (1) “If you could only choose one Barbie to play with, which one you would choose?” and (2) “Is there any Barbie you would not want to play with?” In addition, participants were asked to provide the reasons for their play preferences.

2.2.4. Body dissatisfaction

Body dissatisfaction was assessed via the Children’s Body Image Scale (CBIS; Truby & Paxton, 2010). This pictorial scale contains seven photos of girls representing the standard percentile curves for body mass index. The figures are rated on a continuous scale from 1 to 7 according to increasing body size. Each participant selects the girl that she believes looks most like her (i.e., perceived body size), as well as her ideal figure. Body dissatisfaction was assessed as the difference between the two scores (i.e., perceived body size minus ideal body size). Thus, greater positive scores (i.e., scores of 6 to 1) indicate that the respondent believes she is larger than her ideal; more negative scores (i.e., scores of −6 to −1) indicate that the respondent believes she is smaller than her ideal. A score of zero indicates that the respondent believes her current body matches her ideal. This scale has demonstrated adequate validity and reliability in children as young as 7 years of age (Truby & Paxton, 2010). Therefore, in the current study, the CBIS was only administered to participants age 7 and older.

2.3. Procedure

Participants were recruited through elementary and preschools in southern California. All procedures were approved by the institutional review board at Pepperdine University prior to the start of the study. Parents at participating schools received a letter detailing the study aims and procedures. Interested parents provided written consent allowing their child to participate in the study. Written or verbal assent was obtained from each participant prior to the start of the study. All study procedures took place at the child’s school during their normal school day.

On testing day, each participant was escorted from her class-room to the testing room by a staff member at the school. For all tasks, the child was seated at a comfortable chair and table. An undergraduate research assistant aided the first author in administering the tasks and recorded the child’s responses. The order of the tasks was randomized for each participant. Additionally, the order of the presentation of the Barbie dolls was randomized for each participant at the start of the session. At the end of the session, each child had the opportunity to choose a small prize or sticker.

2.4. Data analyses

Frequencies were utilized to examine the percentage of children who assigned each positive or negative attribute to a given doll, as well as responses to the questions “If you could only choose one Barbie to play with, which one would you choose?” and “Is there any Barbie you would not want to play with?” Participants’ reasons for selecting specific dolls to play with or avoid were coded by a trained research assistant and the frequency of responses was examined. Next, the mean number of positive or negative attributes assigned to each doll was calculated. To handle the count variables, which exhibited non-normal distributions, two Poisson regression analyses with pairwise comparisons were used to examine differences in the number of positive and negative attributes assigned to each doll (e.g., did the dolls vary in the number of positive adjectives assigned to them?). Specifically, doll was entered as a dummy-coded independent variable predicting the number of positive attributes and the number of negative attributes. In addition, four generalized estimating equations with a Poisson distribution were used to examine differences in the number of positive versus negative adjectives used for a specific doll (i.e., were there significantly more positive or negative adjectives attributed to a given doll?). For these analyses, attitude was entered as a dummy-coded independent variable (positive versus negative) predicting the number of attributes assigned to each doll. Generalized linear models with Poisson distributions were used to examine the associations between participant age, their level of body dissatisfaction, their perceived body size, and the number of positive or negative attributes that they assigned to each doll. Finally, in order to examine the impact of Barbie ownership on body dissatisfaction and perceptions of the Fashionista dolls, generalized linear models with Poisson distributions were used to examine differences in body dissatisfaction, as well as the number of positive and negative attributes assigned to each doll between respondents who did own a Barbie and those who did not. Specifically, Barbie ownership was entered as a dichotomous independent variable predicting CBIS body dissatisfaction scores, and the number of positive and negative attributes assigned to each doll. All analyses were conducted using SPSS 25.0.

3. Results

The percentage of respondents identifying each doll as representing a specific attribute is presented in Table 1. With respect to the negative attributes, the curvy doll was most commonly selected for “not pretty” (53.6%) and “has no friends” (42.9%). The original doll was most frequently selected for “sad” (34.5%) and “not smart” (29.8%), while the tall doll was most commonly selected for “mean” (33.3%). The curvy doll was most commonly identified as the doll that participants would not want to play with (39.0%), while the original doll was least commonly selected (1.7%). The most commonly identified reason for not wanting to play with the curvy doll was her larger size, with 25% of the girls describing the doll as “fat” or “chubby,” and 5.6% of the girls describing the doll as “big.” Although the four dolls were presented with identical hair (e.g., color, cut, styling), 14% of girls indicated that they did not like the curvy doll’s hair and 25% of the girls were not able to identify a reason for not wanting to play with the curvy doll.

Table 1.

Percentage of Attributes Assigned to Each Barbie and Doll Selection Preferences.

Original Petite Tall Curvy
Positive attributes
 Happy 22.4% 18.8% 42.4% 16.5%
 Smart 29.8% 22.6% 31.0% 16.7%
 Has friends 21.4% 33.3% 35.7% 9.5%
 Pretty 25.0% 28.6% 36.9% 9.5%
 Helps others 33.3% 22.6% 19.0% 25.0%
Negative attributes
 Sad 34.5% 33.3% 9.5% 22.6%
 Not smart 29.8% 26.2% 16.7% 27.4%
 Has no friends 17.9% 26.2% 13.1% 42.9%
 Not pretty 16.5% 14.3% 16.7% 53.6%
 Mean 19.0% 17.9% 33.3% 29.8%
Choose one doll 21.4% 39.3% 33.3% 6.0%
Not want to play with 1.7% 8.5% 23.7% 39.0%
Open in a new tab

Note. N = 84.

With respect to the positive attributes, the tall doll was most frequently selected for “happy” (42.4%), “smart” (31.0%), “has friends” (35.7%), and “pretty” (36.9%), while the original doll was most frequently selected for “helps others” (33.3%). The curvy doll was the least commonly selected doll for “happy,” “smart,” “has friends,” and “pretty.” Participants most commonly identified the petite doll as the Barbie they would choose to play with (39.3%), while the curvy doll was least commonly chosen (6.0%). Children provided a diverse set of reasons for selecting the petite doll including feeling that the doll looked most similar to themselves (12.1%), liking her small size (9.1%), personality (9.1%), general appearance (9.1%), and short stature (3.0%).

Results from the Poisson regressions indicated significant variation in the number of positive and negative attributes assigned to each doll (see Table 2). Specifically, the tall, original, and petite dolls were all assigned more positive adjectives compared to the curvy doll, while the tall doll received significantly more positive adjectives compared to the petite doll. In addition, the curvy doll was assigned more negative attributes compared to all other dolls.

Table 2.

Poisson Regression Results Comparing the Mean (SD) Number of Positive and Negative Attributes Assigned to Each Barbie.

Original Petite Tall Curvy Wald χ2 (df = 3) p Pairwise Comparisons
Positive attributes 1.32 (1.02) 1.26 (1.03) 1.64 (1.07) 0.77 (0.84) 25.16 < .001 T, O, P > C; T > P
Negative attributes 1.17 (0.94) 1.18 (1.08) 0.89 (0.92) 1.76 (1.23) 26.24 < .001 C > O, P, T
Open in a new tab

Note. O = Original doll, P = Petite doll, T = Tall doll, C = Curvy doll. All pairwise comparisons listed were significant at p ≤ .05. N = 84.

Generalized estimating equations indicated that the tall doll was assigned significantly more positive adjectives than negative adjectives, Wald χ2(1, 84) = 14.13, p < .001, and the curvy doll was assigned significantly more negative adjectives than positive adjectives, Wald χ2(1, 84) = 5.05, p < .001. There were no significant differences between the number of positive and negative adjectives assigned to the original doll, Wald χ2(1, 84) = 0.70, p = .403, or the petite doll, Wald χ2(1,84) = 0.18, p = .675.

Results from the generalized linear models (see Table 3) indicated a significant negative association between body dissatisfaction and the number of negative attributes assigned to the original doll. Specifically, girls who perceived a larger discrepancy between their current size and their desired size, such that they perceived themselves to be larger than their ideal, tended to assign fewer negative attributes to the original doll. Perceived body size demonstrated a negative association with the number of negative attributes assigned to the original doll, and a positive correlation with the number of negative attributes attributed to the petite doll. In other words, girls with a larger perceived body size attributed fewer negative attributes to the original doll and more negative attributes to the petite doll. All other examined associations were not significant.

Table 3.

Associations Among Age, Number of Barbies Owned, Body Dissatisfaction, Perceived Body Size, and Number of Positive or Negative Attributes Assigned to Each Doll.

Original
 Petite
 Tall
 Curvy
Positive
attributes		 Negative
attributes		 Positive
attributes		 Negative
attributes		 Positive
attributes		 Negative
attributes		 Positive
attributes		 Negative
attributes
Age .04 (.04) −.00 (.04) .06 (.04) −.04 (.04) −.06 (.04) −.01 (.05) −.02 (.05) .04 (.03)
Body dissatisfaction .11 (.09) −.24 (.09)* .04 (.08) .10 (.10) −.13 (.09) .04 (.11) −.03 (.12) .09 (.07)
Perceived body size .17 (.11) −.25 (.12)* .03 (.10) .26 (.12)* −.18 (.11) .15 (13) −.03 (.15) −.07 (.09)
Open in a new tab

Note: Associations expressed as unstandardized beta weights and standard errors. Body dissatisfaction and perceived body size was only assessed among participants age seven and older (n = 40). Sample size for analyses examining age was 84.

*

p < .05.

Finally, generalized linear models indicated that the number of positive or negative adjectives assigned to each doll did not differ as a function of Barbie ownership (see Table 4). Body dissatisfaction was also not significantly different between these groups.

Table 4.

Poisson Regression Results Comparing the Number of Positive or Negative Attributes Assigned to Each Doll and Body Dissatisfaction Between Respondents Who Do or Do Not Own a Barbie.

Does own a Barbie (n = 64) Does not own a Barbie (n = 20) Wald χ2 (df = 1) p
Original
 Positive attributes 1.38 (1.03) 1.15 (0.99) 0.58 .45
 Negative attributes 1.11 (0.96) 1.35 (0.88) 0.75 .39
Petite
 Positive attributes 1.16 (0.98) 1.60 (1.14) 2.36 .13
 Negative attributes 1.25 (1.05) 0.95 (1.15) 1.16 .28
Tall
 Positive attributes 1.69 (1.08) 1.50 (1.05) 0.33 .57
 Negative attributes 0.84 (0.88) 1.05 (1.05) 0.72 .40
Curvy
 Positive attributes 0.78 (0.90) 0.75 (0.64) 0.02 .89
 mNegative attributes 1.80 (1.31) 1.65 (0.93) 0.19 .66
Body dissatisfaction 0.30 (0.26) 0.29 (0.57) 0.01 .98
Open in a new tab

Note. Body dissatisfaction was only assessed among participants age seven and older (n = 40). Sample size for analyses examining positive and negative attributes for each doll was 84.

4. Discussion

Consistent with previous literature demonstrating a preference for thinness in young children (Harriger et al., 2010; Musher-Eizenmann et al., 2004), it was hypothesized that girls in the current study would attribute more positive adjectives to the original Barbie doll and more negative adjectives to the curvy Barbie doll. Results supported this hypothesis in comparisons between the two dolls. Girls more frequently selected the original doll as representing all five of the positive attributes (i.e., happy, smart, has friends, pretty, helps others), while the curvy doll was more commonly selected for three of the five negative attributes (i.e., has no friends, not pretty, mean) compared to the original doll. Additionally, girls most commonly identified the curvy doll as the doll that they would not want to play with, while the original doll was least commonly selected for this question. When probed further, the most common reason for not wanting to play with the curvy doll related to her body size. Children referred to the curvy doll as “big,” “fat,” or “chubby,” while the original doll was commonly described as “pretty.” Collectively, these results corroborate extant work demonstrating children’s weight bias against larger individuals (Harriger et al., 2019; Spiel et al., 2012; Worobey & Worobey, 2014) and positive attitudes towards thinner individuals (Harriger et al., 2010).

Given the novelty of the tall and petite sized dolls, no specific hypothesizes were forwarded with regard to participants’ responses to these dolls. Overall, girls in the current study demonstrated the most favorable attitudes towards the tall doll. For example, girls attributed significantly more positive adjectives to the tall doll compared to the petite and curvy Barbie dolls, and fewer negative adjectives to the tall doll compared to the curvy doll. Across all dolls, the tall doll was most likely to be selected for four out of the five positive attributes (i.e., happy, smart, has friends, pretty) and least likely to be selected for three out of the five negative attributes (i.e., sad, not smart, has no friends). In addition, one third of the girls indicated a preference to play with the tall doll. Anecdotal comments made by participants who selected the tall doll indicated that many of them may have assumed that the taller doll was older, perhaps due to her increased “height.” For instance, during the testing session (although not in direct response to prompts regarding doll preference), a number of children spontaneously stated that the taller doll looked like she was more “grown up” or talked about how she must be in a higher grade compared to the other dolls. It is therefore possible that the participants viewed the taller doll as a figure that they would like to emulate when they were older or that they viewed being older as more positive, though future research would be needed to test this hypothesis.

Interestingly, despite demonstrating more favorable attitudes towards the tall Barbie in the adjective attribution task, girls most commonly identified the petite Barbie doll as the doll they would most want to play with. Although the rationale for the selection of the petite doll differed across participants, many reported that they identified most with this doll, or liked her small size or short stature. In addition, children appeared to attribute traits of sociability and kindness to the petite doll. For example, the petite doll was least likely to be associated with meanness, and one third of the sample indicated that the petite doll was most likely to have friends. Therefore, it is possible that participants admired the tall doll but affiliated more with the petite doll, perhaps due to her reduced “height.” Interestingly, petite Barbie is estimated to have the smallest US dress size even when compared to original and tall Barbie (Bates, 2016). Therefore, it is also possible that participants preferred the petite Barbie due to her thin size.

Although previous work examining the impact of age on weight-based attitudes is somewhat mixed (Harriger, 2015; Harrigeret al., 2019; Holub, 2008), we hypothesized that older girls would demonstrate greater preference for thinness (i.e., more positive and fewer negative attributes assigned to the original, tall, and petite dolls) and more negative attitudes towards the curvy doll (i.e., more negative and fewer positive attributes assigned to the curvy doll). However, this hypothesis was not supported. Age was not significantly associated with attitudes towards the tall, curvy, original, or petite dolls.

Finally, it was hypothesized that girls with higher body dissatisfaction would demonstrate greater preference for thinness, manifesting more positive attitudes towards the original doll and more negative attitudes towards the curvy doll. However, this hypothesis was only partially supported. Girls with who perceived their bodies to be larger than their ideal demonstrated less negative attitudes towards the original doll, but body dissatisfaction was not associated with attitudes towards the curvy, petite, or tall dolls. Thus, the current study provides some indication that decreased negative attitudes towards thin bodies is associated with more negative perceptions of one’s own body. This finding is consistent with previous work indicating that body dissatisfaction is related to a preference for thinness (Rodgers, McLean, & Paxton, 2015; Thompson, Heinberg, Altabe, & Tantleff-Dunn, 1999). Importantly, sample size for analyses involving body dissatisfaction was limited (n = 40), as the Children’s Body Image Scale is only validated for children age 7 and older. Therefore, these analyses were underpowered to detect small effects.

Barbie ownership was not related to body dissatisfaction, or the number of positive or negative adjectives assigned to each doll in our sample of young girls. Importantly, however, the majority of the participants in the current study owned a Barbie doll and those who did not own a doll at the time of the study reported previous exposure to Barbie’s. Thus, Barbie ownership does not appear to represent meaningful proxy for Barbie exposure. Rather, given the nearly ubiquitous exposure to Barbie in this sample, results related to Barbie ownership should be interpreted with caution.

Although Mattel does not publish data regarding the sales of the Fashionista doll line, it is notable that none of the participants in the current study reported owning a curvy Barbie doll. Moreover, while some girls (6%) indicated a preference to play with the curvy doll, many girls demonstrated negative attitudes towards this doll, suggesting that they were unlikely to seek out the curvy doll in the future. Thus, although the availability of body diversity in Barbie dolls is an admirable step, these results question the degree to which children will choose to interact with dolls that represent larger (and more realistic) body sizes and shapes. Thus, identifying avenues for increasing the acceptability of larger dolls represents an important step for future work. It is possible that positive marketing and normalization of dolls with larger figures could be used to counteract potentially-harmful weight-based attitudes among youngsters. For example, marketing for the original Barbie highlights her numerous positive traits, which likely contributes to children’s generally positive perception of the doll. Similar marketing campaigns for the Fashionista dolls, which might seek to associate positive characteristics with diverse body sizes may help to counteract negative attitudes towards larger bodies. It is also possible that continued exposure to the new line of dolls may increase familiarity with and positive attitudes towards dolls of varying sizes. Future work should seek to examine whether positive marketing and greater interactions with dolls of diverse body sizes and shapes, such as the curvy Barbie, increases preference for these dolls and positive attitudes towards larger bodies more generally.

This study is the first to examine beliefs about body shape and size in young girls using the Fashionista Barbie dolls, but it is not without limitations. First, the small sample size may have limited our ability to detect significant effects. This was particularly problematic with the analyses that included body dissatisfaction, as only girls that were seven or older completed this measure. Second, the light skin color of the dolls did not enable us to provide all of our participants with stimuli that were consistent with their own skin color. Future researchers may want to consider using dolls with various skin colors in order to assess the interaction between participant race, and the body size and skin color of the doll. Third, participants in this study were all female, and research including boys is warranted as well. Fourth, although prior work has utilized the adjective attribution task to examine weight-based attitudes among youth (e.g., Harriger et al., 2010; Worobey & Worobey, 2014), the nature of the task and the response format (i.e., forced-choice, nominal data) does not allow for participants to rate all figures using an interval or continuous scale, and precludes the ability to examine the psychometric properties of the constructed scales (e.g., negative adjectives). Therefore, future research may consider use of alternative tasks or response scaling to facilitate psychometric investigation of the scales. Future research may also benefit from an assessment of the perceived age of stimulus materials (i.e., figures and dolls) in order to examine whether children associate size with age, as well as the effect that perceived doll age many have on children’s attitudes towards the dolls. Fifth, although we sought to examine relationships between body size attitudes and children’s own perceived body size, future work may seek to utilize objective measurements of children’s body size (e.g., body mass index z-scores). Finally, the cross-sectional design of the study prevents us from making causal conclusions regarding Barbie doll exposure and beliefs about body size and shape. Continued experimental and longitudinal work examining the impact of exposure to diverse dolls on girls’ weight attitudes and body satisfaction is encouraged.

In sum, the results of the current study suggest the presence of potentially harmful attitudes about body size and shape among young girls aged 3 to 10. As exposure to thin idealized dolls may be one means through which these attitudes develop, the creation of curvier commercially-available dolls represents an important step towards combatting thin-ideal internalization and weight bias in youngsters. Indeed, researchers have been calling for such action for over a decade (Dittmar et al., 2006). However, results from the current study suggest that the simple availability of body-diverse dolls may be insufficient to fully combat pervasive cultural messages about body size. We applaud Mattel’s production of more diverse dolls and acknowledge this as a positive initial first step. Moving forward, we encourage the continued pursuit of systematic attempts to reduce weight bias and to encourage acceptance of diverse body sizes and shapes in young children.

References

  1. Anschutz DJ, & Engels RC (2010). The effects of playing with thin dolls on body image and food intake in young girls. Sex Roles, 63, 621–630, https://dx.doi.org/10.1007%2Fs11199-010-9871-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bates C (2016). How does curvy Barbie compare with an average woman? BBC News Magazine,. Retrieved from. http://www.bbc.com/news/magazine-35670446
  3. Boyd H, & Murnen SK (2017). Thin and sexy vs. Muscular and dominant: Prevalence of gendered body ideals in popular dolls and action figures. Body Image, 21, 90–96. 10.1016/j.bodyim.2017.03.003 [DOI] [PubMed] [Google Scholar]
  4. Christel DA, & Dunn SC (2016). Average American women’s clothing size: Comparing national health and nutritional examination surveys (1998–2010) to ASTM International missus and women’s plus size clothing. International Journal of Fashion Design Technology and Education, 10, 129–136. 10.1080/17543266.2016.1214291 [DOI] [Google Scholar]
  5. Cramer P, & Steinwert T (1998). Thin is good, fat is bad: How early does it begin? Journal of Applied Developmental Psychology, 19, 429–451. 10.1016/S0193-3973(99)80049-5 [DOI] [Google Scholar]
  6. Groesz LM, Levine MP, & Murnen SK (2002). The effect of experimental presentation of thin media images on body satisfaction: A meta analytic review. The International Journal of Eating Disorders, 31, 1–16. 10.1002/eat.10005 [DOI] [PubMed] [Google Scholar]
  7. Dittmar H (2012). Dolls and action figures In Cash TF (Ed.), Encyclopedia of body image and human appearance (Vol. 1)(pp. 386–391). San Diego, CA: Elsevier Academic Press. [Google Scholar]
  8. Dittmar H, Halliwell E, & Ive S (2006). Does Barbie make girls want to be thin? The effect of experimental exposure to images of dolls on the body image of 5- to 8-year-old girls. Developmental Psychology, 42, 283–292. 10.1037/0012-1649.42.2.283 [DOI] [PubMed] [Google Scholar]
  9. Dittmar H, & Howard S (2004). Thin-ideal internalization and social comparison tendency as moderators of media models’ impact on women’s body-focused anxiety. Journal of Social and Clinical Psychology, 23, 768–791. 10.1521/jscp.23.6.768.54799 [DOI] [Google Scholar]
  10. Frederick DA, Daniels EA, Bates ME, & Tylka TL (2017). Exposure to thin-ideal media affect most, but not all, women: Results from the perceived effects of media exposure scale and open-ended responses. Body Image, 23, 188–205. 10.1016/j.bodyim.2017.10.006 [DOI] [PubMed] [Google Scholar]
  11. Harriger JA(2015). Age differences in body size stereotyping in a sample of preschool girls. Eating Disorders, 23, 177–190. 10.1080/10640266.2014.964610, http:// [DOI] [PubMed] [Google Scholar]
  12. Harriger JA, Calogero RM,Witherington DC, & Smith JA (2010). Body size stereotyping and internalization of the thin-ideal in preschool-age girls. Sex Roles, 63, 609–620, doi: 10.1007/s11199-010-9868-1. [DOI] [Google Scholar]
  13. Harriger JA, & Thompson JK (2012). Psychological consequences of obesity: Weight bias and body image in overweight and obese youth. International Review of Psychiatry, 24, 247–253. 10.3109/09540261.2012.678817 [DOI] [PubMed] [Google Scholar]
  14. Harriger J, Trammell J, Wick M, & Leudke M (2019). Gender and age differences in preschoolers’ weight bias beliefs and behavioural intentions. The British Journal of Developmental Psychology,. Advance online publication. [DOI] [PubMed] [Google Scholar]
  15. Holub S (2008). Individual differences in the anti-fat attitudes of preschool-children: The importance of perceived body size. Body Image, 5, 317–321. 10.1016/j.bodyim.2008.03.003 [DOI] [PubMed] [Google Scholar]
  16. In Depth: Barbie by the Numbers. (2009). In depth: Barbie by the numbers (March). Retrieved from. https://www.forbes.com/2009/03/05/barbie-design-manufacturing-business_numbers_slide/#110ef9d914a1
  17. Klaczynski P, Daniel DB, & Keller PS (2009). Appearance idealization, body esteem, causal attributions, and ethnic variations in the development of the obesity stereotypes. Journal of Applied Developmental Psychology, 30, 537–551. 10.1016/j.appdev.2008.12.031 [DOI] [Google Scholar]
  18. Kornilaki E (2015). Obesity bias in children: The role of actual and perceived body size. Infant and Child Development, 24,365–378. 10.1002/icd.1894 [DOI] [Google Scholar]
  19. Kuther TL, & McDonald E (2004). Early adolescents’ experiences with, and views of, Barbie. Adolescence, 39, 39–51. [PubMed] [Google Scholar]
  20. Latner JD, Simmonds M, Rosewall JK, & Stunkard AJ (2007). Assessment of obesity stigmatization in children and adolescents. Modernizing a standard measure. Obesity, 15, 3078–3085. 10.1038/oby.2007.366 [DOI] [PubMed] [Google Scholar]
  21. Levine MP, & Harrison K (2004). The role of mass media in the perpetuation and prevention of negative body image and disordered eating In Thompson JK (Ed.), Handbook of eating disorders and obesity (pp. 695–717). New York, NY: Wiley. [Google Scholar]
  22. Lind A, & Brzuzy S (2008). Battleground: Women, gender, and sexuality. Westport, CT: Greenwood Publishing Group. [Google Scholar]
  23. Meers MR, Koball AM, Oehlhof MW, Laurene KR, & Musher-Eizenman DR (2011). Assessing anti-fat bias in preschoolers: A comparison of a computer generated line-drawn figure array and photographic figure array. Body Image, 8, 293–296. 10.1016/j.bodyim.2011.04.006 [DOI] [PubMed] [Google Scholar]
  24. Musher-Eizenman D, Holub S, Miller A, Goldstein SL, & Edwards-Leeper L (2004). Body size stigmatization in preschool children: The role of control attributions. Journal of Pediatric Psychology, 29, 613–620. 10.1093/jpepsy/jsh063 [DOI] [PubMed] [Google Scholar]
  25. Paxton SJ, & Damiano SR (2017). The development of body image and weight bias in childhood In Benson JB (Ed.), Advances in child development and behavior (Vol. 52) (pp. 269–298). Burlington, MA: Academic Press. [DOI] [PubMed] [Google Scholar]
  26. Pedersen EL, & Markee NL (1991). Fashion dolls: Representations of ideals of beauty. Perceptual and Motor Skills, 73, 93–94. [DOI] [PubMed] [Google Scholar]
  27. Pearl RL, & Puhl RM (2014). Measuring internalized weight attitudes across body weight categories: Validation of the modified Weight Bias Internalization Scale. Body Image, 11, 89–92. 10.1016/j.bodyim.2013.09.005 [DOI] [PubMed] [Google Scholar]
  28. Pearl RL, White MA, & Grilo CM (2014a). Overvaluation of shape and weight as a mediator between self-esteem and weight bias internalization among patients with binge eating disorder. Eating Behaviors, 15, 259–261. 10.1016/j.eatbeh.2014.03.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Pearl RL, White MA, & Grilo CM (2014b). Weight bias internalization, depression, and self-reported health among overweight binge eating disorder patients. Obesity, 22, e142–e148. 10.1002/oby.20617 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Puhl RM, & Heuer CA (2009). The stigma of obesity: A review and update. Obesity, 17, 941–964. 10.1038/oby.2008.636 [DOI] [PubMed] [Google Scholar]
  31. Puhl RM, & King KM (2013). Weight discrimination and bullying. Best Practice & Research Clinical Endocrinology & Metabolism, 27, 117–127. 10.1016/j.beem.2012.12.002 [DOI] [PubMed] [Google Scholar]
  32. Rice K, Prichard I, Tiggemann M, & Slater A (2016). Exposure to Barbie: Effects of thin-ideal internalization, body-esteem, and body dissatisfaction among young girls. Body Image, 19, 142–149. 10.1016/j.bodyim.2016.09.005 [DOI] [PubMed] [Google Scholar]
  33. Rodgers RF, McLean SA, & Paxton SJ (2015). Longitudinal relationships among internalization of the media ideal, peer social comparison, and body dissatisfaction: Implications for the tripartite influence model. Developmental Psychology, 51, 706–713. 10.1037/dev0000013 [DOI] [PubMed] [Google Scholar]
  34. Spiel EC, Paxton SH, & Yager Z (2012). Weight attitudes in 3- to 5-year-old children: Age differences and cross sectional predictors. Body Image, 9, 524–527. 10.1016/j.bodyim.2012.07.006 [DOI] [PubMed] [Google Scholar]
  35. Swami V, Frederick DA, Aavik T, Alcalay L, Allik J, Anderson D,… & Zivcic-Becirevic I (2010). The attractive female body weights and body dissatisfaction in 26 countries across 10 world regions: Results from the International Body Project. Personality & Social Psychology Bulletin, 36, 309–325, doi: 0.1177/0146167209359702. [DOI] [PubMed] [Google Scholar]
  36. Suisman JL, O’Conner SM, Sperry S, Thompson JK, Keel PK, Burt SA, … & Klump KL (2012). Genetic and environmental influences on thin-ideal internalization. The International Journal of Eating Disorders, 45, 942–948. 10.1002/eat.22056 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Thompson JK, Heinberg LJ, Altabe MN, & Tantleff-Dunn S (1999). Exacting beauty: Theory, assessment, and treatment of body image disturbance. Washington, DC: American Psychological Association; 10.1037/10312-000 [DOI] [Google Scholar]
  38. Thompson JK, & Stice E (2001). Thin-ideal internalization: Mounting evidence for a new risk factor for body-image disturbance and eating pathology. Current Directions in Psychological Science, 10, 181–183, https://doi.org/10.1111%2F1467-8721.00144. [Google Scholar]
  39. Tiggemann M (2011). Sociocultural perspectives on human appearance and body image In Cash TF & Smolak L (Eds.), Body image: A handbook of science, practice, and prevention (pp. 12–19). New York, NY: Guilford Press. [Google Scholar]
  40. Truby H, & Paxton SJ (2010). Development of the Children’s Body Image Scale. British Journal of Developmental Psychology, 41, 185–203. [DOI] [PubMed] [Google Scholar]
  41. Turkel AR (1998). All about Barbie: Distortions of a transitional object. The Journal of the American Academy of Psychoanalysis, 26, 165–177. [DOI] [PubMed] [Google Scholar]
  42. Turnbull JD, Heaslip S, & McLeod HA (2000). Pre-school children’s attitudes to fat and normal male and female stimulus figures. International Journal of Obesity, 24, 1075–1076. [DOI] [PubMed] [Google Scholar]
  43. Veldhuis J, tel Poel F, Pepping R, Konijn EA, & Spekman MLC (2017). “Skinny is prettier and normal: I want to be normal”—Perceived body image of non-Western ethnic minority children in the Netherlands. Body Image, 20, 74–86. 10.1016/j.bodyim.2016.11.006 [DOI] [PubMed] [Google Scholar]
  44. Vintage Barbie Struts Her Stuff. (2006). Vintage barbie struts her stuff. (September). Retrieved from. http://news.bbc.co.uk/2/hi/business/5370398.stm [Google Scholar]
  45. Winterman D (2009). What would a real life Barbie look like? BBC News Magazine,. Retrieved from. http://news.bbc.co.uk/2/hi/uk_news/magazine/7920962.stm
  46. Worobey J, & Worobey HS (2014). Body-size stigmatization by preschool girls: In a doll’s world, it is good to be “Barbie”. Body Image, 11, 171–174. 10.1016/j.bodyim.2013.12.001 [DOI] [PubMed] [Google Scholar]
  47. Wright D, & Bradbard M (1980). Body build behavioral stereotypes, self-identification, preference and aversion in black preschool children. Perceptual & Motor Skills, 51, 1047–1050. [Google Scholar]
  48. Zuba A, & Warschburger P (2018). Weight bias internalization across weight categories among school-aged children. Validation of the Weight Bias internalization Scale for Children. Body Image, 25, 56–65. 10.1016/j.bodyim.2018.02.008 [DOI] [PubMed] [Google Scholar]

RESOURCES