Psychology of Popular Media

From Filters to Body Positivity: Opposing Social Media Messages and

Adolescent Body Image
Jolien Trekels
Online First Publication, September 19, 2024. https://dx.doi.org/10.1037/ppm0000565

CITATION
Trekels, J. (2024). From filters to body positivity: Opposing social media messages and adolescent body
image.. Psychology of Popular Media. Advance online publication. https://dx.doi.org/10.1037/ppm0000565
 Psychology of Popular Media
© 2024 American Psychological Association
ISSN: 2689-6567 https://doi.org/10.1037/ppm0000565

From Filters to Body Positivity: Opposing Social Media Messages and

Adolescent Body Image
Jolien Trekels
Developmental Social Neuroscience Lab, University of North Carolina at Chapel Hill

Today’s social media landscape is highly complex with adolescents encountering multiple, and sometimes
contradictory, messages simultaneously, leading to a need for a more active perspective on media effects.
The current cross-sectional study aims to take a crucial first step in examining how adolescents’ exposure to
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

contradictory appearance messages relates to their body dissatisfaction. Additionally, the resonance of such
This document is copyrighted by the American Psychological Association or one of its allied publishers.

messages among users is tested as a moderator. In March 2020, 278 adolescents (69.4% girls, Mage = 16.20,
SD = 1.73) completed an online survey. Results show that idealized and counter-idealized content are,
respectively, positively and negatively, related to body dissatisfaction. Body dissatisfaction is lowest when
adolescents indicate to be exposed to both idealized and counter-idealized content in a balanced manner but
increases when idealized content outweighs counter-idealized content, particularly. Interestingly, social
resonance, or the extent to which media messages resonate with the social context (i.e., peer conversations),
does not moderate these relations. Overall, these findings provide an understanding of how adolescents’ social
media consumption patterns can shape their body image, and highlight the need for media effects research to
account for the complexities of adolescents’ media diet.

Public Policy Relevance Statement

In today’s digital age, adolescents navigate a complex social media landscape filled with idealized and
body-positive content. This cross-sectional study finds that a balanced exposure to these content types
correlates with lower levels of body dissatisfaction among adolescents; when idealized content out-
weighs body-positive content, dissatisfaction increases. These findings underscore the value of promot-
ing a more balanced media consumption to foster healthier body image outcomes among young people.

Keywords: social media, different messages, adolescents, body image

Social media have seamlessly integrated into the daily routines of 1968) and the body has been identified as salient in defining oneself
many adolescents; Recent reports reveal that approximately 95% of as a person (Kling et al., 2018). Notably, developmental changes
Flemish adolescents engage with their smartphones on a daily basis such as a changing body composition due to puberty heighten adoles-
(Vanwynsberghe et al., 2022). Decades of research have now estab- cents’ awareness of and concern with physical appearance (Williams &
lished a consensus regarding the adverse effects of social media expo- Currie, 2000). Additionally, adolescents become increasingly con-
sure on users’ body image concerns (for reviews, see Fardouly & cerned with others’ perceptions, leading them to evaluate themselves
Vartanian, 2016; Saiphoo & Vahedi, 2019), identifying adolescents against absolute or relative standards, possibly causing psychological
(ages 12–18) as a particularly susceptible group. One key developmen- distress when there is a discrepancy between their ideal self and actual
tal task during adolescence is to develop a stable identity (Erickson, self (Higgins, 1987). Consequently, adolescents are more prone to the
negative effects of encountering unrealistic and idealized images on
social media, which often trigger upward appearance comparison pro-
cesses and the internalization of this appearance ideal (Saiphoo &
Patrick Sweeney served as action editor. Vahedi, 2019; Vuong et al., 2021).
Jolien Trekels https://orcid.org/0000-0002-4632-441X Partly as a response to these concerning findings, there has been
This research was supported by the Research Foundation Flanders (Grant a notable surge in content advocating for positive body image cir-
12X0820N) when the author was affiliated with the KU Leuven, Department culating on social media platforms. This wave of (textual and
of Communication Sciences, Media Psychology Lab. The author has no visual) body positivity messages actively challenges societal
conflicts of interest to declare. This study was not preregistered. The data appearance standards and promotes the cultivation of self-love
that support the findings of the current study are not available to share, but for one’s body (Cohen et al., 2019). For example, images of plus-
analysis code is available upon request to the author. size models with hashtags like #plussizefashion frequently strive to
Jolien Trekels served as lead for conceptualization, data curation, formal
enhance the visibility and normalization of underrepresented body
analysis, funding acquisition, investigation, methodology, project adminis-
tration, resources, software, supervision, and writing–original draft.
types (Cohen et al., 2019). Moreover, body-positive interventions
Correspondence concerning this article should be addressed to Jolien seek to raise users’ awareness of the unrealistic nature of idealized
Trekels, Developmental Social Neuroscience Lab, University of North images online (Gordon et al., 2020). Recent research has begun to
Carolina at Chapel Hill, 235 East Cameron Avenue, Chapel Hill, NC delve into the connection between exposure to such affirming
27599, United States. Email: jolient@unc.edu social media content and its potential impact on body image.

1
 2 TREKELS

Studies by Cohen et al. (2019) and Rodgers et al. (2021) suggest a Hypothesis 1: When exposure to idealized social media content
promising positive effect associated with the consumption of such outweighs exposure to counter-ideal content, there is a positive
content and young women’s body image. It has been argued that association with body dissatisfaction.
body-positive content might reduce users’ engagement in upward
appearance comparisons, leading to decreased negative body On the other end of the continuum, interpretations may be predom-
image and increased positive body image (Rousseau, 2024). inantly informed by counter-idealized contents. Individuals may
Additionally, experimental studies have confirmed that exposure actively engage with and prioritize counter-ideal messages, which
to body diversity images (diverse in ethnicity and body size) can may catch their attention and outweigh the idealization. This suggests
increase body satisfaction and foster negative attitudes toward the that individuals may critically analyze and question the idealized con-
thin ideal (Ogden et al., 2020). tent presented in media, and may interpret it in light of the counter-ideal
It follows, then, that social media present a nearly dual narrative messages they encounter. These counter-messages, which can be both
on body image. Adolescents find themselves immersed in both the visual (e.g., depiction of body diversity, Ogden et al., 2020) and textual
portrayal of idealized and narrowly defined beauty standards and a (e.g., Cohen et al., 2019), have been found to mitigate potential nega-
counter-narrative featuring body-positive content that champions tive effects and promote greater body satisfaction (Rousseau, 2024).
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Consequently, the second hypothesis reads:

This document is copyrighted by the American Psychological Association or one of its allied publishers.

acceptance and diversity simultaneously. This duality gives rise to

a highly intricate media environment, one that adolescents navigate Hypothesis 2: When exposure to counter-idealized social media
while in the process of shaping their body image and self-concept. content outweighs exposure to counter-ideal content, there is a
Studies investigating how adolescents’ exposure to this dual narra- negative association with body dissatisfaction.
tive relates to their body image have been lacking. One recent
study showed that adolescents’ contradictory perceptions of attrac- In between these two extremes, individuals may encounter and pro-
tiveness are curvilinearly related to appearance anxiety (Trekels & cess both idealizing and counter-messages equally, resulting in a
Eggermont, 2024). The current study contributes to this literature balanced state where the idealized and counter-idealized messages can-
by examining how the ratio/balance between idealized and counter- cel each other out. This may lead to a null finding in terms of media
idealized exposure on social media is linked with adolescents’ body effects, as individuals may interpret and process media messages in
dissatisfaction. a way that balances out the ideal and counter-ideal contents, resulting
in no significant effects on their attitudes, beliefs, or behaviors.
The Interplay of Ideal and Counter-Ideal Appearance Hypothesis 3: When individuals encounter idealized and
Messages counter-idealized messages equally, there will be no significant
Considering a broader perspective on the types of appearance association with body dissatisfaction.
messages necessitates assigning a more active role to the viewer.
Drawing from an interpretative approach, such as the message inter- Social Resonance as a Moderator
pretation process model (Austin & Meili, 1994) and media practice
(MP) model (Steele & Brown, 1995), media exposure and process- By framing media processing as a negotiation between different
ing can be considered a conscious yet automated meaning-making types of messages, resulting in a continuum of interpretative imbal-
practice. Consequently, viewers are not passive recipients of ances, this study seeks to provide a more nuanced understanding of
appearance-related messages but are instead cocreators of meaning, how social media relate to individuals’ body dissatisfaction. Drawing
dynamically shaping their self-concept and body image through from the message interpretation process (Austin & Meili, 1994) and
their interactions with media. Active audience perspectives assume MP (Steele & Brown, 1995) models, the current study emphasizes
that users interpret a multitude of messages through their own that body dissatisfaction can fluctuate based on the dynamic interaction
lenses, actively engaging with media content by selecting, interpret- between idealized and counter-idealized messages, influenced by how
ing, and integrating messages based on their personal experiences, individuals actively engage with and interpret media content.
social contexts, and preexisting beliefs. In this study, we specifically Notably, in order to understand which messages might be more
propose a negotiation process between different types of messages, salient to certain adolescents, it is crucial to incorporate the resonance
resulting in varying degrees of interpretative (im)balances between between messages and the individual interpretative lens as it deter-
idealized and counter-idealized messages. The outcome of this mines how information is perceived, processed, interpreted, and inte-
negotiation phase in media processing is a complex continuum, grated into one’s self-concept. The concept of resonance is also
which can be operationalized into three realities. On one end of found in other theoretical perspectives, including cultivation theory
the continuum, the idealized content may overshadow any counter- (Shrum, 2017) and the MP model (Steele & Brown, 1995), which
idealized messages that are present, aligning with the presumption posit that the level of resonance between media messages and the indi-
in most traditional behavioral media effects research that media pri- vidual’s interpretative lens shapes their media effects. In the context of
marily depict behaviors or appearances in a positive or idealized body image development, the peer group may act as a particularly
context (Fischer et al., 2011). In this case, individuals may interpret strong context for adolescents to shape their interpretative lenses of
and process media messages in a way that emphasizes the idealized media messages (Steele & Brown, 1995). Previous research has sug-
content, while downplaying or ignoring the counter-idealized mes- gested that both offline (Jones et al., 2004; Trekels & Eggermont,
sages. Considering prior research has associated exposure to ideal- 2017) and online (Trekels et al., 2018) appearance-related conversa-
ized social media content with increased body dissatisfaction tions among peers can heighten body image concerns. Such discus-
(Fardouly & Vartanian, 2016; Saiphoo & Vahedi, 2019), the first sions may underscore the significance of conforming to appearance
hypothesis reads: standards, thereby intensifying the salience of messages related to
 OPPOSING ONLINE APPEARANCE MESSAGES AND BODY IMAGE 3

these ideals (Clark & Tiggemann, 2007). More frequent engagement ranging from 1 = never to 6 = constantly. The 11 items were based
in these types of conversations has been associated with body dissatis- on prior literature on the narrow appearance standards for men and
faction (Zimmer-Gembeck et al., 2023). Overall, girls have been found women (Thompson & Stice, 2001; Vandenbosch, 2017) and showed
to report more appearance conversations with friends, which may good internal consistency (α = .954). An example item is “On social
amplify the influence of such experiences for girls (Zimmer- media, how often do you see people who are thin, tall, young.”
Gembeck et al., 2023). Still, research has pointed to the importance
of including both girls and boys in body image research (McCabe & Counter-Idealized Exposure on Social Media
Ricciardelli, 2003). In light of these considerations, the current study
aims to investigate whether social interactions centered around appear- Participants also indicated how often they were confronted with
ance moderate the association between different messages encountered counter-idealized content. Based on existing research on body-positive
on social media and girls’ and boys’ body dissatisfaction. content, the measure was divided into two components. The first com-
ponent focused on visual exposure, where participants indicated the fre-
Hypothesis 4: Social resonance (i.e., engagement in a peer quency of seeing body-positive images that depict a wide range of body
appearance culture) moderates the examined association between sizes and appearances (e.g., “On social media, how often do you see
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

exposure to (counter)idealized social media content and body dis- people who are fat, short, not muscled”; 11 items; exploratory factor
This document is copyrighted by the American Psychological Association or one of its allied publishers.

satisfaction, such that those engaging in an appearance culture analysis showed one factor with eigenvalue 5.638, explained variance
might be more prone to outweigh idealized content and experience 56.376%; α = .911). The second component centered on textual
more body dissatisfaction. counter-ideal references. These items were derived from content anal-
ysis studies (e.g., Cohen et al., 2019; Gordon et al., 2020) showing that
The Current Study body-positive content often includes messages that broadly conceptu-
alize beauty (e.g., “messages that tell you that beauty is subjective
To empirically examine the propositions of the current study, a and can be defined differently by each individual”), statements that
cross-sectional survey study was conducted focusing on adolescents’ highlight the editing and alterations required to achieve the appearance
body dissatisfaction and their exposure to appearance-related content of models and famous people (e.g., “messages that highlight the editing
on social media, given their widespread popularity among adoles- and alterations required to achieve the appearance of models
cents (Best et al., 2014) and the documented coexistence of opposing and famous people”) and hashtags on body positivity (e.g., “love
appearance messages (Cohen et al., 2019; Holland & Tiggemann, your body”). A total of nine items were included in the measure.
2016). To capture the intricate dynamics of multilayered media expo- Exploratory factor analysis showed one factor with eigenvalue 5.664,
sure, I employed polynomial regression and response surface analy- explained variance 62.937%. The measure showed good internal con-
sis, allowing for the examination of curvilinear relationships. This sistency with α = .930.
statistical approach enabled me to test how the complex interplay
between ideal and counter-ideal messages in media exposure influ- Body Dissatisfaction
ences adolescents’ body dissatisfaction, taking into account the
nuanced nature of media effects. By applying these advanced statis- Participants’ body dissatisfaction was measured by the Appearance
tical techniques, I aimed to gain a more comprehensive understand- subscale of the Body Esteem Scale (Mendelson et al., 2001). On a
ing of the relationship between media exposure and body 5-point Likert scale, ranging from 1 = never to 5 = always, participants
dissatisfaction in adolescents, considering the potential nonlinearities responded to 10 items about their appearance. Example items are “I
involved in this process. Given that prior research highlights differ- worry about my looks” and “I’m pretty happy about the way I look”
ences in the influence of social media use on body image among (reverse coded). The scale showed good internal consistency (α = .925).
females and males (Rodgers & Rousseau, 2022), I will explore gen-
der differences in the examined relations. Peer Appearance Culture

Method Participants were asked about their engagement in an appearance

culture to assess social resonance. Five statements probed adolescents
In March 2020, adolescent boys and girls participated in an online about how often they discussed appearance-related topics among their
(due to the COVID-19 pandemic) cross-sectional survey study. friends (Jones et al., 2004). An example item is “My friends and I talk
Participants were approached through high schools that agreed to partic- about what we would like our bodies to look like” and adolescents
ipate in the study. Passive parental consent and adolescent assent were responded on a 5-point scale, ranging from 1 = never to 5 = very
obtained prior to participation; the ethical protocol was approved by often. The scale showed good internal consistency (α = .882).
KU Leuven’s ethical board (Sociaal-Maatschappelijke Ethische
Commissie). A total of 306 adolescents (69.4% girls) completed the sur- Covariates
vey; 28 participants were omitted because of missing values (at least
50%) on the key variables in this study. The analytical sample thus con- Covariates in the current study included body mass index (BMI) cal-
sisted of 278 adolescents (69.42% girls, Mage = 16.20, SD = 1.73, age culated from self-reported weight and height, age, and biological sex
range = 12–19, 98.9% indicated to be born in Belgium). (1 = boy, 2 = girl).

Idealized Exposure on Social Media Data Analysis Plan

To assess adolescents’ exposure to idealized messages on social In order to fully understand the interplay of exposure to idealized
media, they were asked to rate 11 items on a 6-point Likert scale, and (textual or visual) counter-idealized content on social media and
 4 TREKELS

how it relates to adolescents’ body dissatisfaction, polynomial counter-idealized content was significantly linked with their TikTok
regression and response surface analyses were conducted use (Table 2).
(Edwards & Parry, 1993). To alleviate the threats of multicollinear-
ity between the component measures and their higher order terms,
the exposure measures were centered around the pooled grand
Polynomial Regression and Surface Response Analysis
mean (Aiken et al., 1991). Then, two separate polynomial regression Idealized Versus Visual Counter-Idealized Content on
analyses were conducted—one including ideal exposure versus tex- Social Media
tual counter-idealized exposure and one including ideal exposure
versus visual counter-idealized exposure. Body dissatisfaction was The first step in polynomial regression is a linear regression analysis,
regressed on control variables (i.e., BMI, gender, and age), centered without the interaction between the standardized predictor variables
idealized exposure (I ), centered visual/textual counter-idealized and the squared standardized predictor variables. This linear regression
exposure (C ), idealized exposure squared (I 2), ideal times counter- analysis (controlling for BMI, gender, and, age) indicated that both ide-
ideal exposure (I × C ), and counter-idealized exposure squared alized exposure (β = .152, t = 2.475, p = .014) and visual counter-
(C 2). Next, we analyzed these models separately for boys and girls. idealized content (β = −.177, t = −2.891, p = .004) were sig-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

nificantly related to body dissatisfaction (R = .397, R 2 = .157),

This document is copyrighted by the American Psychological Association or one of its allied publishers.

In addition, two moderated polynomial regression analyses were

conducted to investigate whether social resonance (i.e., peer appearance F(5, 235) = 8.776, p , .001.
culture) would moderate the curvilinear relationships between ideal Next, a hierarchical polynomial regression and surface response
versus (textual or visual) counter-ideal exposure and body dissatisfac- analysis of the interaction between idealized and visual counter-
tion. To be specific, in the first step, control variables were entered. idealized social media content on body dissatisfaction was performed
Then, I − C, I + C, IC were entered. In the third step, appearance cul- (controlling for BMI, gender, and age). The results indicated that R 2
ture (W ) was entered. And finally, I × W, C × W, I 2 × W, C 2 × W, and did only marginally significantly increase when adding the product
I × C × W were entered. of the standardized predictor variables and the squared standardized
predictor variables into the regression equation (p = .086); as such
results should be interpreted cautiously. However, the squared stan-
Results dardized visual counter-idealized exposure reached significance.
Descriptive Statistics and Preliminary Analyses Therefore, I performed a response surface analysis and graphed the
results (see Figure 1). This analysis revealed that the surface under
Descriptive statistics and zero-order correlations were run for all var- the incongruence line (I = −C) curved downward (curvature, B =
iables (Table 1). Girls and boys differed significantly on all key vari- 0.18, SE = 0.05, p , .001). Figure 1 indicates that there is a
ables, except for counter-idealized visual exposure on social media. U-shaped line along the incongruence line (i.e., the line from the left
We also correlated our measures on idealized and counter-idealized front corner to the right rear corner). The curvature along the I = −C
exposure on social media with participants’ reported frequency of line indicates that body dissatisfaction is lowest when early adolescents’
social media use. Participants were asked to report how often exposure to idealized and visual counter-idealized content is aligned,
they used social media apps (Instagram, Facebook, Snapchat, and and any deviation from the congruence line—thus moving more to
TikTok) on an 8-point Likert scale (1 = less than 10 min or never to its right or left—increases body dissatisfaction. In addition, the slope
8 = more than 5 hr). These results showed that idealized content was of the incongruence line (slope, B = −0.28, SE = 0.07, p , .001) is
significantly linked with their reported frequency of using Instagram, negative, indicating that body dissatisfaction is higher when the dis-
Snapchat, and TikTok. Participants’ reported exposure to visual crepancy is such that idealized exposure is higher than visual counter-

Table 1
Descriptive Statistics and Zero-Order Correlations
Variable 1 2 3 4 5 6 7 8
1. Idealized social media exposure — .054 −.162*** .186** .236*** .002 −.026 .156**
2. Textual counter idealized social media exposure — .453*** −.067 .138* −.022 −.012 .243***
3. Visual counter idealized social media content — −.147* .107 −.070 .096 .105
4. Body dissatisfaction — .392*** .195** .137* .254***
5. Social resonance — .145* .155** .388***
6. BMI — .402*** .128*
7. Age — .170**
8. Gender —
M (SD) 4.616 2.942 2.535 2.627 2.417 20.171 16.200
(0.882) (1.003) (0.689) (0.823) (0.944) (2.945) (1.727)
Mgirls (SD) 4.708 3.106 2.584 2.766 2.659 20.427 16.395
(0.851) (1.046) (0.710) (0.836) (0.935) (2.886) (1.629)
Mboys (SD) 4.412 2.581 2.428 2.313 1.864 19.621 15.759
(0.851) (0.792) (0.631) (0.707) (0.707) (3.012) (1.865)
Significant difference between boys and girls * * * * * *
Note. BMI = body mass index.
* p , .05. ** p , .01. *** p , .001.
 OPPOSING ONLINE APPEARANCE MESSAGES AND BODY IMAGE 5

Table 2 idealized social media content on body dissatisfaction was performed

Zero-Order Correlations Between Idealized and Counter-Idealized (controlling for BMI, gender, and age). The results indicated that R2
Social Media Content and Frequency of Social Media did only marginally significantly increase when adding the product of
the standardized predictor variables and the squared standardized pre-
Exposure variables Instagram Facebook Snapchat TikTok
dictor variables into the regression equation (p = .063), as such the
Idealized exposure on social .199*** .034 .211*** .128* results should be interpreted cautiously. However, the standardized
media idealized exposure and the interaction term reached significance.
†
Counter-idealized exposure .103 −.072 .018 .125*
on social media—visual Therefore, I performed a response surface analysis and graphed the
Counter-idealized exposure .079 −.064 −.019 .076 results (see Figure 2). This analysis revealed that the surface under
on social media—textual the incongruence line (X = −Y ) curved downward (curvature, B =
†
p , .10. * p , .05. *** p , .001. 0.25, SE = 0.09, p , .001). Figure 2 indicates that there is a
U-shaped line along the incongruence line (i.e., the line from the
left front corner to the right rear corner). The curvature along the
X = −Y line indicates that body dissatisfaction is lowest when early
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

idealized exposure (rather than vice versa). This effect is visualized in

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Figure 1, in which body dissatisfaction is lower at the right front corner adolescents’ exposure to idealized and textual counter-idealized con-
(where I = −2 and C = 2) than at the right rear corner (where I = 2 and tent is aligned, and any deviation from the congruence line–thus
C = −2). moving more to its right or left increases body dissatisfaction. In
addition, the slope of the incongruence line (slope, B = −0.26,
Idealized Versus Textual Counter-Idealized Content on SE = 0.09, p , .001) is negative, indicating that body dissatisfaction
Social Media is higher when the discrepancy is such that idealized exposure is
higher than textual counter-idealized exposure (rather than vice
The linear regression analysis, without the interaction between the versa). This effect is visualized in Figure 2, in which body dissatisfac-
standardized predictor variables and the squared standardized predic- tion is lower at the right front corner (where X = −2 and Y = 2) than
tor variables, indicated that only idealized exposure (β = .181, t = at the right rear corner (where X = 2 and Y = −2).
2.740, p = .007) and not textual counter-idealized content (β =
−.130, t = −1.839, p = .067) significantly related with body dissat- Gender Differences
isfaction (R = .385, R 2 = .148), F(5, 235) = 8.156, p , .001.
Next, a hierarchical polynomial regression and surface response First, measurement invariance tests were performed to establish
analysis of the interaction between idealized and visual counter- whether the meaning of the latent variables was the same in both

Figure 1
Results of the Response Surface Analysis for Body Dissatisfaction as Predicted by Idealized and
Visual Counter-Idealized Social Media Exposure

Curvilinear association between idealized versus visual counter-idealized social

media exposure and body dissatisfaction

Z 3
(Body Dissatisfaction
2

0
2 C
1 0 -1 (visual counter idealized
I -2 -2
(Idealized social media exposure)
social media exposure)

Note. See the online article for the color version of this figure.
 6 TREKELS

Figure 2
Results of the Response Surface Analysis for Body Dissatisfaction as Predicted by Idealized and
Textual Counter-Idealized Social Media Exposure

Curvilinear association between idealized versus textual counter-idealized

social media exposure and body dissatisfaction

Z 3
(Body Dissatisfaction)
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

0
X
2 1 0 (textual Counter ideal
-1 -2 -2
Y social media exposure)
(Idealized
social media exposure)

Note. See the online article for the color version of this figure.

groups (Mplus, Version 8.7). Fit indices supported configural, met- these results again warrant caution. Specifically, the response surface
ric, and scalar invariance for body dissatisfaction and exposure to analysis showed only a significant slope (B = −0.31, SE = 0.09,
idealized social media content. The difference in fit between the con- p = .003) not curvature of the incongruence line (B = −0.24, SE =
figural and metric models was nonsignificant, body dissatisfaction, 0.01, p = .107) which means that as the difference between idealized
χ2(9) = 8.526, p = .482; idealized exposure, χ2(10) = 14.959, and counter-idealized textual content increases, body dissatisfaction
p = .134. The overall fit for scalar invariance was good; compared decreases in a linear manner.
with the metric model, the chi-square test showed variance, body For boys, the R 2 values of both polynomial regressions were insig-
dissatisfaction, χ2(9) = 11.179, p = .264; idealized exposure, nificant (for idealized and counter-idealized visual content;
χ2(10) = 17.991, p = .055. Metric and configural invariance was R 2 = .056, p = .689; for idealized and counter-idealized textual con-
established for exposure to counter-idealized visual and textual tent; R 2 = .048, p = .571). As such, we did not perform a response
social media content, with a nonsignificant difference between the surface analyses were not performed.
metric and configural models, textual, χ2(8) = 10.597, p = .256;
visual, χ2(9) = 2.792, p = .972. Scalar invariance could not be Moderated Polynomial Regression and Response Surface
established for the counter-idealized exposure measures, textual,
Analysis
χ2(8) = 29.627, p , .001; visual, χ2(9) = 50.948, p , .001.
Next, the polynomial regressions were tested separately for girls The result of this analysis revealed that there was only a significant
and boys. For girls, I found a significant polynomial regression for direct effect of appearance culture on body dissatisfaction (β = .330,
idealized and counter-idealized visual social media content on body t = 5.132, p , .001). However, there was no significant moderation
dissatisfaction (R 2 = .176, p = .011). The response surface analysis effect of appearance culture on the curvilinear relationships between ide-
showed that the surface under the incongruence line (I = −C) curved alized versus visual counter-idealized exposure (ΔR2 = .268, ΔF2 =
downward (curvature, B = 0.28, SE = 0.10, p , .01), indicating that 0.199, p = .962), nor for the curvilinear relationship between idealized
girls’ body dissatisfaction is lowest when their exposure to idealized versus textual counter-idealized exposure (ΔR2 = .523, ΔF2 = 0.857,
and visual counter-idealized content is aligned, and any deviation p = .510). Response surfaces were therefore not examined.
from the congruence line—thus moving more to its right or left—
increases body dissatisfaction. In addition, the slope of the incongru- Discussion
ence line (slope, B = −0.32, SE = 0.09, p , .001) is negative, indi-
cating that body dissatisfaction is higher when the discrepancy is This study serves as a significant initial step toward recognizing the
such that idealized exposure is higher than visual counter-idealized importance of considering the diversity of media messages, particularly
exposure (rather than vice versa). For the association between ideal- those related to appearance. Participants indicated to frequently encoun-
ized and counter-idealized textual social media content, we found a ter idealized content on social media but they also indicated to encoun-
marginally significant regression (R 2 = .145, p = .087). Therefore, ter counter-idealized content to some extent. The current findings
 OPPOSING ONLINE APPEARANCE MESSAGES AND BODY IMAGE 7

thereby align with recent studies indicating the complexity of users’ More research is needed to confirm the current study’s findings
experiences with idealized content with some engaging in protective given its cross-sectional design that precludes any causal inferences.
filtering (Evens et al., 2021). Notably, our study adds to this literature Additionally, despite the premise of audience research (e.g., Austin
by emphasizing that both types of media exposure, ideal and counter- & Meili, 1994) that media consumption and reception is an activity
ideal, are intricately intertwined and should be considered concurrently embedded within the interpersonal context (Moores, 1993), the moder-
as factors shaping adolescent body image. ating role of social resonance was not supported in the current study. It
Both the observed curvilinear relationships between media expo- is possible that our measurement approach, which did not specifically
sure and body dissatisfaction and the association between different differentiate between idealized and counter-idealized appearances in
types of content and body dissatisfaction in opposing directions friends’ conversations, may have contributed to this null finding.
(i.e., idealized and counter-idealized content were, respectively, pos- Future studies are encouraged to further unravel the different “lived
itively and negatively related to body dissatisfaction) are in line with experiences” that shape interpretative lenses (e.g., prior dispositions;
the hypotheses. Notably, users appear to be (a) navigating a complex due to the cross-sectional nature of the current study, we could not accu-
and diverse media landscape where (b) not all messages are equally rately differentiate state from trait variables). Assessing pubertal status
seen or processed by the audience. In particular, the results of our might be another important line for future studies as pubertal status has
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

study revealed that body dissatisfaction was lowest when adolescents been linked with adolescent body image (de Guzman & Nishina,
were exposed to both idealized and counter-idealized content in a 2014). Moreover, incorporating experimental eye-tracking studies
balanced manner, but increased—particularly—when idealized con- could provide valuable insights into whether the observed curvilinear
tent outweighed counter-idealized content. In other words, this find- relationships are a result of not seeing or not processing the information
ing underscores the significance of considering the intricate interplay presented in media messages. Additionally, it should be noted that the
between different messages, as it reflects the variable vulnerability of results from the response surface analysis should be interpreted with
adolescents to the social media effect. Understanding this dynamic caution considering the R2 increase was only marginally significant.
can serve as a valuable guide in the development of effective inter- However, because the current study is the first to systematically include
ventions aimed at promoting healthy body image among adolescents. exposure to different messages into one model, we believe it is of
In line with previous research, our findings indicated a positive exploratory nature and therefore valuable. Future studies are encour-
correlation between exposure to idealized content and the use of aged to continue this line of work.
Instagram, Snapchat, and TikTok among participants (Tylka et al., By considering the interplay between different and opposing media
2023). Additionally, I observed that exposure to visual counter- messages, as well as the role that viewers themselves bring to the media
idealized content was correlated with adolescents’ use of TikTok, experience, the current study took a first step in accounting for the com-
suggesting that TikTok might also serve as a platform for encounter- plexities of the contemporary media landscape. Further research in this
ing more diverse appearance-related content (Seekis & Lawrence, area has the potential to enhance our understanding of the nuanced
2023). ways in which social media influence individuals’ perceptions and
Interestingly, the results suggest a potential difference between behaviors in relation to body image and other relevant outcomes.
textual and visual references to body positivity in their association
with body dissatisfaction. Specifically, while exposure to idealized
content was positively associated with body dissatisfaction, only References
visual counter-idealized content was negatively associated with Aiken, L. S., West, S. G., & Reno, R. R. (1991). Multiple regression: Testing
body dissatisfaction. This finding aligns with studies indicating and interpreting interactions. Sage Publication.
that images are particularly attention grabbing (McManis, et al., Austin, E. W., & Meili, H. K. (1994). Effects of interpretations of televised
2001) and may be more persuasive by eliciting stronger emotional alcohol portrayals on children’s alcohol beliefs. Journal of Broadcasting
responses (Powell et al., 2015). In the context of this study, body- & Electronic Media, 38(4), 417–435. https://doi.org/10.1080/0883815940
positive images might thus elicit stronger responses and more signif- 9364276
icantly influence adolescents’ feelings toward their bodies. Best, P., Manktelow, R., & Taylor, B. (2014). Online communication, social
At the same time, I did observe that a balance between idealized con- media and adolescent wellbeing: A systematic narrative review. Children
and Youth Services Review, 41, 27–36. https://doi.org/10.1016/j.childyouth
tent and both textual and visual counter-idealized content was associ-
.2014.03.001
ated with the lowest levels of body dissatisfaction. Therefore, studies Clark, L., & Tiggemann, M. (2007). Sociocultural influences and body image
incorporating both types of content are warranted, as they could provide in 9 to 12-year-old girls: The role of appearance schemas. Journal of
crucial insights into how different social media platforms—whether Clinical Child & Adolescent Psychology, 36(1), 76–86. https://doi.org/
more text based or visual based—relate to positive body image. 10.1080/15374410709336570
The results further support prior research suggesting that girls may Cohen, R., Irwin, L., Newton-John, T., & Slater, A. (2019). #Bodypositivity:
be more strongly influenced by their exposure to both idealized and A content analysis of body positive accounts on Instagram. Body Image,
counter-idealized content (Rodgers & Rousseau, 2022). Specifically, 29, 47–57. https://doi.org/10.1016/j.bodyim.2019.02.007
the findings showed that girls reported more frequent exposure to de Guzman, N. S., & Nishina, A. (2014). A longitudinal study of body dis-
satisfaction and pubertal timing in an ethnically diverse adolescent sam-
appearance-related content—both idealized and counter-idealized—
ple. Body Image, 11(1), 68–71. https://doi.org/10.1016/j.bodyim.2013
on social media compared to boys. Additionally, I found stronger evi-
.11.001
dence for the importance of a balance between these two types of con- Edwards, J. R., & Parry, M. E. (1993). On the use of polynomial regression
tent in relation to girls’ body dissatisfaction. It appeared that, especially equations as an alternative to difference scores in organizational research.
among girls, an imbalance between these two increases body Academy of Management Journal, 36(6), 1577–1613. https://doi.org/10
dissatisfaction. However, given the lower sample size of boys in our .2307/256822
study, these findings should be replicated in future research. Erickson, E. H. (1968). Identity youth and crisis. Norton.
 8 TREKELS

Evens, O., Stutterheim, S. E., & Alleva, J. M. (2021). Protective filtering: A Saiphoo, A. N., & Vahedi, Z. (2019). A meta-analytic review of the relationship
qualitative study on the cognitive strategies young women use to promote between social media use and body image disturbance. Computers in Human
positive body image in the face of beauty-ideal imagery on Instagram. Behavior, 101, 259–275. https://doi.org/10.1016/j.chb.2019.07.028
Body Image, 39, 40–52. https://doi.org/10.1016/j.bodyim.2021.06.002 Seekis, V., & Lawrence, R. K. (2023). How exposure to body neutrality content
Fardouly, J., & Vartanian, L. R. (2016). Social media and body image concerns: on TikTok affects young women’s body image and mood. Body Image, 47,
Current research and future directions. Current Opinion in Psychology, 9, Article 101629. https://doi.org/10.1016/j.bodyim.2023.101629
1–5. https://doi.org/10.1016/j.copsyc.2015.09.005 Shrum, L. J. (2017). Cultivation theory: Effects and underlying processes. In
Fischer, P., Greitemeyer, T., Kastenmüller, A., Vogrincic, C., & Sauer, A. P. Rössler, C. A. Hoffner, & L. van Zoomen (Eds.), The international
(2011). The effects of risk-glorifying media exposure on risk-positive cog- encyclopedia of media effects (pp. 1–12). Wiley-Blackwell. https://
nitions, emotions, and behaviors: A meta-analytic review. Psychological doi.org/10.1002/9781118783764
Bulletin, 137(3), 367–390. https://doi.org/10.1037/a0022267 Steele, J. R., & Brown, J. D. (1995). Adolescent room culture:
Gordon, C. S., Rodgers, R. F., Slater, A. E., McLean, S. A., Jarman, H. K., & Studying media in the context of everyday life. Journal of Youth and
Paxton, S. J. (2020). A cluster randomized controlled trial of the SoMe
Adolescence, 24(5), 551–576. https://doi.org/10.1007/BF01537056
social media literacy body image and wellbeing program for adolescent
Thompson, J. K., & Stice, E. (2001). Thin-ideal internalization: Mounting
boys and girls: Study protocol. Body Image, 33, 27–37. https://doi.org/
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

evidence for a new risk factor for body-image disturbance and eating
10.1016/j.bodyim.2020.02.003
This document is copyrighted by the American Psychological Association or one of its allied publishers.

pathology. Current Directions in Psychological Science, 10(5), 181–183.

Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect.
https://doi.org/10.1111/1467-8721.00144
Psychological Review, 94(3), 319–340. https://doi.org/10.1037/0033-295X
Trekels, J., & Eggermont, S. (2017). Beauty is good: The appearance culture,
.94.3.319
Holland, G., & Tiggemann, M. (2016). A systematic review of the impact of the the internalization of appearance ideals, and dysfunctional appearance beliefs
use of social networking sites on body image and disordered eating out- among tweens. Human Communication Research, 43(2), 173–192. https://
comes. Body Image, 17, 100–110. https://doi.org/10.1016/j.bodyim.2016 doi.org/10.1111/hcre.12100
.02.008 Trekels, J., & Eggermont, S. (2024). Adolescents’ multi-layered media pro-
Jones, D. C., Vigfusdottir, T. H., & Lee, Y. (2004). Body image and the appear- cessing: A panel study on positive and negative perceptions toward ideals
ance culture among adolescent girls and boys: An examination of friend and adolescents’ appearance anxiety. Communication Research, 51(6),
conversations, peer criticism, appearance magazines, and the internalization 607–630. https://doi.org/10.1177/00936502231151471
of appearance ideals. Journal of Adolescent Research, 19(3), 323–339. Trekels, J., Ward, L. M., & Eggermont, S. (2018). I “like” the way you look:
https://doi.org/10.1177/0743558403258847 How appearance-focused and overall Facebook use contribute to adoles-
Kling, J., Wängqvist, M., & Frisén, A. (2018). “This body is me”: Discovering cents’ self-sexualization. Computers in Human Behavior, 81, 198–208.
the ways in which the body is salient in people’s identities. Body Image, 24, https://doi.org/10.1016/j.chb.2017.12.020
102–110. https://doi.org/10.1016/j.bodyim.2017.12.009 Tylka, T. L., Rodgers, R. F., Calogero, R. M., Thompson, J. K., & Harriger, J.
McCabe, M. P., & Ricciardelli, L. A. (2003). Sociocultural influences on body A. (2023). Integrating social media variables as predictors, mediators, and
image and body changes among adolescent boys and girls. The Journal of moderators within body image frameworks: Potential mechanisms of
Social Psychology, 143(1), 5–26. https://doi.org/10.1080/00224540309 action to consider in future research. Body Image, 44, 197–221. https://
598428 doi.org/10.1016/j.bodyim.2023.01.004
McMANIS, M. H., Bradley, M. M., Berg, W. K., Cuthbert, B. N., & Lang, P. Vandenbosch, L. P. (2017). Media representation: Health and body images.
J. (2001). Emotional reactions in children: Verbal, physiological, and In P. Roessler (Ed.), The international Encyclopedia of media effects (pp.
behavioral responses to affective pictures. Psychophysiology, 38(2), 1–13). Wiley. https://doi.org/10.1002/9781118783764.wbieme0145
222–231. https://doi.org/10.1111/1469-8986.3820222 Vanwynsberghe, H., Joris, G., Waeterloos, C., Anrijs, S., Vanden Abeele,
Mendelson, B. K., Mendelson, M. J., & White, D. R. (2001). Body-esteem M., Ponnet, K., De Wolf, R., Van Ouytsel, J., Van Damme, K.,
scale for adolescents and adults. Journal of Personality Assessment, Vissenberg, J., d’Haenens, L., Zenner, E., Peters, E., De Pauw, S.,
76(1), 90–106. https://doi.org/10.1207/S15327752JPA7601_6 Frissen, L., & Schreuer, C. (2022). Onderzoeksrapport Apestaartjaren:
Moores, S. (1993). Interpreting audiences: The ethnography of media con-
De digital leefwereld van kinderen en jongeren [Research report apestaart-
sumption. SAGE.
jaren: The digital world of children and adolescents]. Mediaraven.
Ogden, J., Gosling, C., Hazelwood, M., & Atkins, E. (2020). Exposure to body
Vuong, A. T., Jarman, H. K., Doley, J. R., & McLean, S. A. (2021). Social
diversity images as a buffer against the thin-ideal: An experimental study.
media use and body dissatisfaction in adolescents: The moderating role of
Psychology, Health & Medicine, 25(10), 1165–1178. https://doi.org/10
thin- and muscular-ideal internalisation. International Journal
.1080/13548506.2020.1734219
Powell, T. E., Boomgaarden, H. G., De Swert, K., & de Vreese, C. H. (2015). of Environmental Research and Public Health, 18(24), Article 13222.
A clearer picture: The contribution of visuals and text to framing effects. https://doi.org/10.3390/ijerph182413222
Journal of Communication, 65(6), 997–1017. https://doi.org/10.1111/ Williams, J. M., & Currie, C. (2000). Self-esteem and physical development
jcom.12184 in early adolescence: Pubertal timing and body image. The Journal of
Rodgers, R. F., Paxton, S. J., & Wertheim, E. H. (2021). #Take idealized bod- Early Adolescence, 20(2), 129–149. https://doi.org/10.1177/02724316000
ies out of the picture: A scoping review of social media content aiming to 20002002
protect and promote positive body image. Body Image, 38, 10–36. https:// Zimmer-Gembeck, M. J., Hawes, T., Scott, R. A., Campbell, T., & Webb, H.
doi.org/10.1016/j.bodyim.2021.03.009 J. (2023). Adolescents’ online appearance preoccupation: A 5-year longi-
Rodgers, R. F., & Rousseau, A. (2022). Social media and body image: tudinal study of the influence of peers, parents, beliefs, and disordered eat-
Modulating effects of social identities and user characteristics. Body ing. Computers in Human Behavior, 140, Article 107569. https://doi.org/
Image, 41, 284–291. https://doi.org/10.1016/j.bodyim.2022.02.009 10.1016/j.chb.2022.107569
Rousseau, A. (2024). Body-positive Instagram exposure and young women’s
body image: The mediating role of appearance comparison and broadly Received November 17, 2023
conceptualizing beauty. Health Communication, 39(8), 1520–1531. Revision received June 28, 2024
https://doi.org/10.1080/10410236.2023.2222460 Accepted August 19, 2024 ▪