Table of Contents
Type: Full Undergraduate Dissertation | Subject: Psychology | Level: Undergraduate | Word Count: ~9,300 words
This model Full Undergraduate Dissertation was produced by an Essays UK specialist as reference material for learning purposes only. For support in this field, see our psychology dissertation writers.
Overview: This model BSc Psychology dissertation presents a complete, structurally faithful example of a cross-sectional correlational study examining the relationship between social media use and body image among UK undergraduate women. It demonstrates how a full final-year project is assembled — abstract, introduction, literature review, methodology, results, discussion and conclusion — with illustrative data, formal hypothesis testing, structured statistical reporting and Harvard (Cite Them Right) referencing. All figures, sample characteristics and statistics are illustrative and generated for teaching purposes; the document is intended solely as a learning reference for how an undergraduate psychology dissertation is organised, argued and reported.
Background: Concern about the influence of social media on young women’s body image has grown rapidly, yet much public debate focuses on the sheer quantity of use rather than how platforms are used. Theory suggests that appearance-focused activity and the social comparisons it invites may be more consequential for body image than total screen time.
Aim: To examine the relationships between overall social media use, appearance-focused social media use, appearance-based social comparison, thin-ideal internalisation and body dissatisfaction in a sample of UK undergraduate women, and to establish which variables best account for body dissatisfaction.
Method: A cross-sectional correlational design was used. A convenience sample of 187 women aged 18–25 studying at UK universities completed an online questionnaire comprising a measure of daily social media use, an appearance-focused use scale, the Physical Appearance Comparison Scale-Revised, the thin-ideal internalisation subscale of the Sociocultural Attitudes Towards Appearance Questionnaire-4, and the Body Shape Questionnaire. Data were analysed using Pearson correlations and hierarchical multiple regression. All data are illustrative and generated for this model dissertation.
Results: Body dissatisfaction was weakly associated with daily time on social media (r = .21) but moderately to strongly associated with appearance-focused use (r = .46), appearance comparison (r = .58) and thin-ideal internalisation (r = .52). In the regression model, appearance comparison, thin-ideal internalisation and appearance-focused use each predicted body dissatisfaction, together explaining 41% of the variance, whereas daily time ceased to be a significant predictor once these variables were entered.
Conclusion: The illustrative findings support the view that the nature of social media use, rather than its quantity, is most relevant to body image. The results are consistent with social comparison and sociocultural theories and point towards comparison and internalisation as appropriate targets for media-literacy interventions. Limitations of the cross-sectional, self-report design and directions for future research are discussed.
Body image — the multidimensional set of perceptions, thoughts and feelings a person holds about their physical appearance — has long been recognised as central to psychological wellbeing, particularly among young women. Negative body image, and in particular body dissatisfaction, is a robust and well-replicated correlate of depressed mood, lowered self-esteem, disordered eating and reduced quality of life. Because body dissatisfaction is so prevalent among women in Western cultures that researchers have described it as normative, understanding the social and environmental factors that sustain it remains an important task for psychology.
Over the past two decades, the media environment within which young women form and evaluate their body image has changed profoundly. Where earlier generations of research concentrated on traditional mass media such as television, magazines and advertising, the contemporary landscape is dominated by social networking platforms. Image-centred platforms in particular have reorganised how appearance is presented, viewed and judged. Content is no longer produced solely by professional media organisations; it is generated continuously by peers, acquaintances and influencers, is highly visual, is frequently edited or filtered, and is accompanied by quantified social feedback in the form of likes, comments and follower counts. Young women are simultaneously the audience for, and the producers of, this appearance-saturated content.
These features have prompted widespread concern that social media may be especially corrosive to young women’s body image. The argument is intuitively compelling: if exposure to idealised images in traditional media was associated with body dissatisfaction, then a medium that delivers a near-constant stream of curated, idealised and interactive appearance content should be at least as influential, and perhaps more so. However, intuition is not evidence, and the empirical picture is more nuanced than public discussion sometimes implies. A central distinction that has emerged in the literature is between the quantity of social media use and its quality or nature. Studies that measure only total time spent online tend to report weaker and less consistent associations with body image than studies that measure the specific activities users engage in, such as viewing and posting appearance-related content, or the psychological processes that use activates, such as social comparison.
This distinction matters because it bears directly on how society responds to the perceived risks of social media. If harm were a simple function of exposure, then the appropriate response would be to limit access or reduce time online, and much popular commentary and some policy proposals adopt precisely this logic. If, on the other hand, harm arises chiefly from particular activities and from the psychological processes they set in motion, then the more promising response is to change how platforms are used and to equip young people to engage with appearance content more critically. The two accounts carry different implications for parents, educators, universities, clinicians and platform designers, and distinguishing between them empirically is therefore not merely an academic exercise but a prerequisite for sound intervention. The present study was designed with this distinction at its centre.
Despite a rapidly expanding evidence base, several gaps justify further investigation. First, much of the foundational research was conducted on adolescent samples or on undergraduates in North America and Australia. The UK undergraduate population is an important and somewhat distinct group: students are typically in the developmental transition of emerging adulthood, are heavy and sophisticated users of social media, and are navigating a period of identity consolidation and heightened social evaluation. Comparatively less work has focused specifically on UK undergraduate women.
Second, studies vary considerably in how they operationalise social media use, and these differences may explain inconsistencies in the literature. A design that measures overall use, appearance-focused use, the comparison process and thin-ideal internalisation within a single sample is well placed to clarify which constructs carry the association with body dissatisfaction, and in particular to test whether the nature of use outperforms its quantity. Establishing this has practical importance: advice to simply reduce screen time rests on different assumptions from advice to change how platforms are used or to build resilience to comparison.
Third, the present project is designed as a transparent, replicable, theory-driven example of a correlational undergraduate dissertation. It therefore foregrounds the links between theory, hypotheses, measurement and analysis, and reports its findings in a structured manner consistent with the conventions of psychological science.
The overarching aim of the study was to examine the relationship between social media use and body image among UK undergraduate women, distinguishing the contribution of the quantity of use from that of appearance-focused use and the associated psychological processes. The specific objectives were:
The study addressed two research questions. First, how are overall social media use, appearance-focused use, appearance comparison and thin-ideal internalisation related to body dissatisfaction in UK undergraduate women? Second, which of these variables best accounts for body dissatisfaction, and does the nature of use predict body dissatisfaction beyond the quantity of use? On the basis of the theoretical and empirical literature reviewed in Chapter 2, five directional hypotheses were advanced:
By modelling the relative contributions of quantity and quality of social media use within a single UK undergraduate sample, the study speaks to a question of both theoretical and applied relevance. Theoretically, confirming that comparison and internalisation outperform sheer time would support process-oriented accounts of media influence rather than simple exposure accounts. Practically, such a finding would suggest that interventions should prioritise reshaping how young women engage with appearance content and reducing the tendency to compare, rather than focusing narrowly on limiting time online. Although the data in this model dissertation are illustrative, the design demonstrates how such questions can be addressed rigorously at undergraduate level.
The study is also significant for the groups most directly concerned. For young women themselves, a clearer understanding of mechanism supports a more empowering message than blanket warnings about screen time: it suggests that how one curates a feed and whether one engages in constant comparison are modifiable behaviours, rather than implying that the technology is uniformly and unavoidably harmful. For universities and student-wellbeing services, identifying comparison and internalisation as key correlates indicates where limited resources might most usefully be directed. For researchers, the design illustrates the value of measuring the specific activities and processes implicated by theory rather than relying on undifferentiated indices of use. In these respects the project aspires to practical as well as scholarly relevance, notwithstanding the illustrative status of its data.
The remainder of the dissertation is organised as follows. Chapter 2 reviews the conceptual and empirical literature on body image and social media, develops the theoretical framework and identifies the gap the study addresses. Chapter 3 sets out the methodology, including the research design, participants, measures, procedure, ethical considerations and analysis strategy. Chapter 4 reports the results of the data screening, descriptive statistics, correlational analysis and regression. Chapter 5 discusses the findings in relation to the hypotheses and existing literature, considers theoretical and practical implications, and acknowledges limitations. Chapter 6 concludes and summarises the contribution of the work.
Body image is widely understood as a multifaceted construct encompassing perceptual, cognitive, affective and behavioural components. The perceptual component concerns the accuracy with which individuals estimate their own body size and shape; the cognitive component concerns the thoughts, beliefs and investment attached to appearance; the affective component concerns feelings such as satisfaction, anxiety or shame; and the behavioural component concerns actions such as body checking, avoidance or appearance management. Cash (2004) argued that body image is best conceived not as a static trait but as a dynamic set of attitudes and experiences that are activated and shaped by context, a view that is particularly pertinent to the study of social media, where appearance-relevant cues are encountered continuously and in highly social settings.
Within this broad construct, body dissatisfaction — a negative subjective evaluation of one’s physical body, frequently focused on weight and shape — has received the most empirical attention because of its links to clinically significant outcomes. Body dissatisfaction is a prospective risk factor for the development of disordered eating and is associated with depression, anxiety and low self-esteem. Its high prevalence among women has led some scholars to describe it as a normative discontent, underlining the importance of understanding the sociocultural forces that cultivate and maintain it. The present study treats body dissatisfaction as the principal outcome variable, operationalised through concern about body shape.
It is important to note that body image is not simply the absence of pathology, and contemporary scholarship has increasingly distinguished negative body image from positive body image — a construct encompassing body appreciation, acceptance and respect that is not merely the opposite pole of dissatisfaction. Although the present study focuses on body dissatisfaction, the recognition that appearance attitudes can be positive as well as negative has practical significance, since it implies that interventions might aim not only to reduce dissatisfaction but to cultivate appreciation of the body for its function and capacities. The construct’s situational responsiveness, emphasised by Cash (2004), also justifies attention to the everyday environments — now including social media feeds — in which appearance attitudes are repeatedly activated, rehearsed and consolidated. A woman’s body image on any given day may be shaped less by stable traits than by the cumulative weight of the appearance-relevant cues and comparisons she has recently encountered.
Three complementary theoretical frameworks inform research on media and body image and together underpin the hypotheses of this study.
Festinger’s (1954) social comparison theory proposes that people possess a fundamental drive to evaluate their own abilities and attributes and, in the absence of objective standards, do so by comparing themselves with others. Comparisons may be lateral, downward (with those seen as worse off) or upward (with those seen as better off). Upward appearance comparisons are of particular relevance to body image: when individuals compare their own appearance unfavourably with others perceived as more attractive, the discrepancy tends to produce dissatisfaction with the self. Social media platforms are a potent engine for such comparisons because they present a dense, continuous and self-selected stream of peers and public figures, frequently depicted at their most attractive, inviting repeated upward comparison. Importantly, the theory locates the harmful effect not in exposure as such but in the comparison process it elicits, which implies that individual differences in the tendency to compare should be an important moderator and that appearance comparison should mediate the relationship between appearance-focused use and body dissatisfaction.
Fredrickson and Roberts (1997) proposed objectification theory to explain how living in a culture that sexually objectifies the female body shapes women’s experiences and mental health. The theory argues that girls and women are socialised to internalise an observer’s perspective on their own bodies, a process termed self-objectification, in which the self is habitually monitored and evaluated on the basis of appearance. Self-objectification is theorised to produce appearance anxiety and body shame and to contribute to a range of mental health risks. Social media extends the opportunities for objectification in novel ways: the self is quite literally presented as an image to be viewed and rated by an audience, and the quantified feedback of likes and comments attaches explicit social value to appearance. The framework thus suggests that appearance-focused engagement with social media may heighten body surveillance and shame, reinforcing dissatisfaction.
Thompson and colleagues (1999) advanced the tripartite influence model, which specifies three primary sociocultural sources of appearance pressure — peers, family and the media — and proposes that their influence on body dissatisfaction and disordered eating operates through two mediating mechanisms: appearance comparison and internalisation of the thin ideal. Thin-ideal internalisation refers to the extent to which an individual cognitively accepts and endorses socially defined ideals of thinness and attractiveness as personal standards against which to evaluate themselves. Thompson and Stice (2001) assembled converging evidence that internalisation functions as a causal risk factor for body-image disturbance and eating pathology. The model is especially useful for the present study because it integrates media influence with the twin processes of comparison and internalisation, and because it predicts that these processes, rather than mere exposure, carry the effect on body dissatisfaction — a prediction that directly motivates H5.
Empirical research on social media and body image has grown rapidly and has been synthesised in several reviews. Holland and Tiggemann (2016) conducted a systematic review of studies linking social networking site use to body image and disordered eating and concluded that use is broadly associated with body-image concern, but that the strength of association depends heavily on how use is measured and on the appearance-related nature of the activities involved. Fardouly and Vartanian (2016), reviewing the field, similarly emphasised that appearance-focused activities and the comparisons they generate are more consistently implicated than global indices of use.
A recurrent theme is the comparatively weak and inconsistent relationship between total time spent on social media and body dissatisfaction. Where associations are found, they tend to be small, and some studies report no significant relationship once appearance-focused variables are controlled. This pattern is theoretically coherent: time spent scrolling through non-appearance content — news, hobbies, messaging friends — would not be expected to activate appearance comparison or threaten body image. It is the appearance-relevant content and the comparison process that theory identifies as harmful, not screen time per se.
The reliance of much early research on cross-sectional designs and single global measures of use has been a persistent methodological criticism of the field. Global self-reported use is a coarse index that aggregates heterogeneous activities — messaging close friends, reading news, watching entertainment and viewing appearance content — most of which have no obvious bearing on body image. Averaging across these activities is likely to dilute any association, which may partly explain why studies relying on such measures report weak and inconsistent effects. A further complication is that the relationship between use and body image is plausibly bidirectional: women who are already dissatisfied with their bodies may be drawn towards appearance-focused content and comparison, so that use both shapes and is shaped by body image. These considerations have prompted calls for more precise measurement of appearance-relevant activity and for designs capable of disentangling the direction of effects, and they reinforce the rationale for the present study’s simultaneous modelling of quantity and quality of use.
Consistent with this reasoning, studies that isolate appearance-focused use report stronger associations with body image. Cohen, Newton-John and Slater (2017) examined appearance-focused activities on Facebook and Instagram and found that such activities, rather than overall use, were associated with greater body image concern in young women, with appearance comparison playing a central role. Experimental work points in the same direction: Fardouly and colleagues (2015) found that brief exposure to Facebook was associated with more negative mood and, among women prone to appearance comparison, greater desire to change their appearance relative to a control condition, indicating that the platform’s effect operates through comparison rather than through exposure alone.
Research also highlights the specific risk posed by particular kinds of content. Tiggemann and Zaccardo (2015) demonstrated that viewing so-called fitspiration imagery — content ostensibly promoting fitness but frequently depicting a narrow, idealised body type — produced poorer body image and mood in young women compared with neutral travel imagery, despite its apparently positive framing. Such findings underscore that it is the idealised, appearance-focused character of content, and the comparisons it invites, that matter, rather than the amount of time spent viewing it. Perloff (2014) drew these threads together in an influential theoretical article, arguing that the interactive, user-generated and feedback-rich features of social media make it a distinctive and potent context for appearance comparison, and calling for research that models the mediating processes rather than treating use as a monolithic exposure variable.
A related and important observation is that the source of the idealised image appears to matter. Some evidence suggests that comparisons with peers and acquaintances encountered on social media may be at least as consequential as comparisons with celebrities or models, perhaps because peers are perceived as more similar and therefore more relevant comparison targets, and because their apparently ordinary yet carefully curated self-presentations can seem more attainable and thus more invidious. The blurring of the boundary between media figures and peers is a distinctive feature of social media: a follower may compare herself in quick succession with a professional model, a well-known influencer and a former schoolmate, all presenting idealised images within the same feed. This density and variety of comparison targets, combined with the self-selected nature of the content, distinguishes social media from the more passive and professionally mediated exposure characteristic of traditional media, and lends theoretical weight to the expectation that appearance comparison will be a particularly powerful correlate of body dissatisfaction in this context.
A further strand of research suggests that effects are not uniform across platforms or activities. Image-centred platforms, where the primary mode of communication is visual and appearance is foregrounded, are generally theorised to pose greater risk than text-centred platforms. Tiggemann and Slater (2013), in an early study of adolescent girls, found that internet exposure, and Facebook use specifically, were associated with greater body image concern, drive for thinness and internalisation, whereas total internet time was less predictive — again suggesting that the appearance relevance of the platform and activity is pivotal. More recent commentary has extended these arguments to highly visual, filter-laden platforms, where editing tools and algorithmic curation may intensify exposure to idealised images and the pressure to present an idealised self.
Taken together, this literature converges on a process account: social media influences body image chiefly by providing abundant opportunities for upward appearance comparison and by reinforcing internalisation of appearance ideals, with appearance-focused activity acting as the gateway to these processes. The quantity of use is, at best, a distal and weak correlate. This account directly motivates the present study’s attempt to pit quantity against quality within a single model.
Although the broad sociocultural account applies to women in general, the strength of the relationship between social media and body image varies considerably between individuals, and a complete model must accommodate these differences. Perhaps the most consistently identified moderator is the trait tendency to engage in appearance comparison. Experimental work indicates that the adverse effects of exposure to idealised images are concentrated among women high in this tendency; those who do not habitually compare appear relatively insulated even when exposed to the same content (Fardouly et al., 2015). This observation is theoretically important because it locates vulnerability not in the medium but in the interaction between the medium and the individual’s characteristic processing of appearance information, and it is one reason the present study measured the comparison tendency directly rather than inferring it from exposure.
Self-esteem is a second relevant individual difference. Low global self-esteem is associated both with a greater propensity to engage in upward social comparison and with heightened sensitivity to its negative consequences, so that women with fragile self-regard may be doubly exposed — more likely to compare and more affected when they do. Thin-ideal internalisation, treated here as a predictor, can also be understood as an individual difference in vulnerability: the more a woman has adopted a narrow appearance ideal as a personal standard, the larger and more distressing any self-ideal discrepancy revealed by comparison is likely to be.
On the protective side, there is growing interest in whether certain forms of engagement might be benign or even beneficial. Exposure to body-positive and appearance-diverse content has been associated in some studies with more favourable body-image outcomes, at least in the short term, suggesting that the effects of social media are not uniformly negative and depend on the content curated and the dispositions brought to it. Media literacy — the capacity to recognise the constructed, edited and commercially motivated nature of appearance content — has similarly been proposed as a buffer. Taken together, this literature indicates that the relationship between social media and body image is conditional rather than universal, shaped by what users view, how they process it and the psychological resources they bring. The present study does not test all of these moderators, but its focus on comparison and internalisation reflects their centrality to this conditional account.
Emerging adulthood, the developmental period spanning roughly the late teens and early twenties, is characterised by identity exploration, instability and heightened sensitivity to social evaluation. University entry typically coincides with this period and introduces new social comparison contexts, new peer networks and, frequently, intensive social media use for social connection. UK undergraduate women are therefore a theoretically interesting population: they are developmentally primed for social comparison, are embedded in appearance-conscious peer cultures, and are among the heaviest and most fluent users of image-centred platforms. Yet much of the influential evidence derives from adolescent samples or from non-UK undergraduates, and the degree to which the quantity-versus-quality pattern holds specifically for UK undergraduate women has received limited direct attention. Examining this population therefore has both theoretical and applied value, not least because universities are well placed to deliver body-image and media-literacy interventions.
The reviewed literature supports three conclusions. First, body dissatisfaction is a prevalent and consequential problem for young women, and sociocultural influences are central to its maintenance. Second, social media is widely implicated, but the association with body dissatisfaction is carried by appearance-focused use and the processes of appearance comparison and thin-ideal internalisation, rather than by the quantity of use. Third, these relationships have been examined less often in UK undergraduate women specifically, and relatively few studies model quantity and quality of use simultaneously to establish their relative contributions. The present study addresses this gap by measuring overall use, appearance-focused use, appearance comparison, thin-ideal internalisation and body dissatisfaction in one UK undergraduate sample, and by testing whether the quality of use predicts body dissatisfaction beyond its quantity.
The study was conducted within a post-positivist philosophical framework. Post-positivism retains the positivist commitment to objective measurement, hypothesis testing and the identification of regularities, while acknowledging that measurement in psychology is fallible, that constructs are theory-laden, and that knowledge is probabilistic rather than certain. This stance is appropriate to the research questions, which concern the existence, direction and relative magnitude of relationships between measurable constructs. Accordingly, a quantitative, deductive approach was adopted: hypotheses were derived from established theory and tested against numerical data using inferential statistics.
A cross-sectional correlational design was used, in which all variables were measured at a single time point via self-report questionnaire and the relationships between them examined without experimental manipulation. A correlational design was appropriate because the research questions concern naturally occurring associations between constructs that cannot ethically or practically be manipulated — it would be neither feasible nor ethical to assign young women to differing levels of appearance comparison or thin-ideal internalisation. The design permits the estimation of the strength and direction of associations and, through multiple regression, the assessment of the unique contribution of each predictor while controlling for the others.
The principal limitation of this design is that it does not permit causal inference: an association between appearance comparison and body dissatisfaction is consistent with comparison causing dissatisfaction, with dissatisfaction increasing the tendency to compare, or with a third variable influencing both. This limitation is acknowledged openly in the reporting and discussion and informs the recommendations for future longitudinal and experimental work. A cross-sectional design was nonetheless judged suitable for an undergraduate project with constrained time and resources, and is well established as a means of mapping relationships that can subsequently be probed causally.
The decision to use hierarchical multiple regression, rather than a series of separate correlations alone, reflects the study’s central comparative question. Entering overall use first and the quality-of-use variables second allows the incremental, unique variance attributable to appearance-focused use, comparison and internalisation to be quantified while statistically holding quantity of use constant. This analytic structure operationalises the theoretical argument directly: if the nature of use matters more than its quantity, then the quality variables should add substantial explained variance at the second step, and the contribution of quantity should diminish once they are entered. A design and analysis that can produce this specific pattern of results, or fail to, is more informative than one that simply establishes that each variable is correlated with the outcome in isolation.
Participants were 187 women aged between 18 and 25 years (M = 20.4, SD = 1.7) who were enrolled as undergraduate students at universities in the United Kingdom. A non-probability convenience and snowball sampling strategy was used, with recruitment conducted through university participant pools, student social media groups and word of mouth. The inclusion criteria were: identifying as a woman; being aged 18–25; being a current undergraduate student at a UK institution; and being an active user of at least one social media platform. Participants who did not meet these criteria, or who did not complete the key measures, were excluded. A total of 194 responses were received and 7 were removed during screening (see Section 4.1), yielding the final analytic sample of 187.
An a priori power analysis indicated that, to detect a medium effect (f² = 0.15) in a multiple regression with four predictors at α = .05 and power of .80, a minimum of approximately 85 participants was required. The achieved sample of 187 therefore provided ample statistical power for the planned analyses. The sample characteristics are summarised in Table 1. (All sample details are illustrative and generated for this model dissertation.)
| Characteristic | Category | n | % |
|---|---|---|---|
| Year of study | First year | 71 | 38.0 |
| Second year | 63 | 33.7 | |
| Third year or above | 53 | 28.3 | |
| Most-used platform | Image-centred platform | 109 | 58.3 |
| Short-video platform | 52 | 27.8 | |
| Text-centred platform | 26 | 13.9 | |
| Daily use | Under 2 hours | 41 | 21.9 |
| 2 to 4 hours | 88 | 47.1 | |
| Over 4 hours | 58 | 31.0 |
The online questionnaire comprised a demographic section followed by five established or adapted self-report measures. All multi-item scales were presented in a fixed order, and item wording was retained from the published instruments where applicable. The measures are summarised in Table 2.
Overall use was assessed with two items capturing self-reported average daily time spent on social media across all platforms, rated on an ordinal scale, and the number of platforms used actively. For the correlational and regression analyses, a continuous estimate of daily minutes was derived from the time item. This global index was intended to represent the quantity of use, independent of its appearance relevance.
Appearance-focused use was measured with an eight-item scale adapted for this study from published appearance-related activity measures. Items assessed the frequency of activities such as viewing and posting photographs of oneself and others, following appearance- and fitness-oriented accounts, editing or filtering one’s own images, and attending to likes and comments on appearance-related posts. Items were rated on a five-point frequency scale, with higher mean scores indicating more appearance-focused engagement. The scale demonstrated good internal consistency in the present sample (α = .82).
The tendency to engage in appearance comparison was measured using the Physical Appearance Comparison Scale-Revised (PACS-R), an 11-item instrument assessing how frequently individuals compare their overall appearance and specific features with those of others. Items were rated on a five-point scale and averaged, with higher scores indicating a greater tendency to compare. The scale showed excellent internal consistency (α = .94).
Thin-ideal internalisation was assessed with the internalisation (thin/low body fat) subscale of the Sociocultural Attitudes Towards Appearance Questionnaire-4 (SATAQ-4). The subscale measures the degree to which respondents endorse a thin, toned appearance ideal as a personal standard. Items were rated on a five-point agreement scale and averaged, with higher scores indicating greater internalisation. Internal consistency was good (α = .89).
Body dissatisfaction, the outcome variable, was measured using the Body Shape Questionnaire (BSQ), a widely used self-report instrument assessing concern about body shape, particularly the experience of feeling fat, over the preceding four weeks. Items were rated on a six-point frequency scale and summed, with higher scores indicating greater body shape concern and dissatisfaction. The scale demonstrated excellent internal consistency (α = .95).
| Construct | Instrument | Items | Response scale | α |
|---|---|---|---|---|
| Overall use | Daily time / platform count | 2 | Ordinal / count | — |
| Appearance-focused use | Adapted activity scale | 8 | 5-point frequency | .82 |
| Appearance comparison | PACS-R | 11 | 5-point frequency | .94 |
| Thin-ideal internalisation | SATAQ-4 (internalisation) | 5 | 5-point agreement | .89 |
| Body dissatisfaction | BSQ | 16 | 6-point frequency | .95 |
The study was administered as an online survey hosted on a secure survey platform. Potential participants accessed the survey via an anonymous link circulated through the recruitment channels described above. On opening the survey, participants were presented with an information sheet detailing the purpose of the study, what participation involved, the voluntary and anonymous nature of participation, their right to withdraw, data handling arrangements and sources of support should any questions cause distress. Participants then provided informed consent by confirming that they had read the information and agreed to take part before any study items were displayed. Eligibility screening questions were presented first; those who did not meet the inclusion criteria were routed to the end of the survey.
Eligible participants completed the demographic questions followed by the five measures. Completion took approximately 12 to 15 minutes. On finishing, participants were shown a debrief screen reiterating the aims of the study, providing fuller detail on the constructs under investigation, and listing support services and helplines relevant to body image and mental health. Contact details for the researcher and supervisor were provided for any subsequent questions or requests to withdraw data. No personally identifying information was collected, and responses were stored securely in anonymised form.
Prior to full data collection, the questionnaire was piloted with a small group of volunteers drawn from the target population to check the clarity of the instructions and items, the functioning of the survey routing and the estimated completion time. The pilot indicated that the measures were well understood and that completion time was acceptable; minor wording adjustments were made to two demographic items for clarity, and no changes were required to the validated scales. Piloting in this way is a straightforward means of improving face validity and reducing the risk of measurement error arising from ambiguous instructions or technical faults, and is good practice in survey-based research.
The study was designed and conducted in accordance with the British Psychological Society’s Code of Ethics and Conduct (2021) and received approval from the departmental research ethics committee prior to data collection. Several ethical issues were given particular attention given the sensitive nature of the topic.
Informed consent and the right to withdraw. Participation was entirely voluntary. Participants gave informed consent after reading a clear information sheet and were advised that they could withdraw at any point during the survey without penalty simply by closing the browser, and could request the removal of their data for a defined period afterwards by quoting a self-generated code.
Anonymity and confidentiality. No names, email addresses or other identifying information were collected. Data were stored on password-protected, encrypted university systems accessible only to the research team, and were reported solely in aggregate form, in line with data protection requirements.
Protection from harm. Because questions about body image and appearance can be distressing, the risk of harm was minimised by using validated, non-intrusive measures, by clearly signposting the topic in advance so that participants could make an informed choice, and by providing a thorough debrief with details of relevant support services and helplines. Participants were reminded that they were free to stop at any time.
Avoidance of deception. No deception was used; the aims of the study were described honestly, with sufficient detail reserved for the debrief only to avoid priming responses, consistent with accepted practice.
Data were analysed using standard statistical software. Prior to hypothesis testing, the data were screened for missing values, out-of-range responses and univariate and multivariate outliers, and the assumptions underlying the planned analyses — normality, linearity, homoscedasticity and the absence of problematic multicollinearity — were examined. Descriptive statistics and internal consistency reliabilities were computed for all scales. Hypotheses 1 to 4 were tested using Pearson product-moment correlations between each predictor and body dissatisfaction. Hypothesis 5 was tested using hierarchical multiple regression, in which overall daily use was entered at the first step and appearance-focused use, appearance comparison and thin-ideal internalisation were entered at the second step, allowing the incremental variance explained by the quality-of-use variables to be assessed. An alpha level of .05 was adopted throughout, and effect sizes are reported alongside significance tests to aid interpretation.
Of the 194 responses received, 4 were removed because the respondents did not meet the inclusion criteria on the screening items, and a further 3 were removed owing to substantial missing data on one or more key measures, leaving a final analytic sample of 187. Among retained cases, missing values on individual items were minimal (under 2% per scale) and were addressed by prorating scale scores where participants had completed at least 80% of a scale’s items. Inspection of standardised scores and boxplots identified a small number of univariate outliers, which were retained after winsorising to the nearest non-outlying value, as their removal did not materially alter the results. Examination of skewness and kurtosis indicated that all scale scores were approximately normally distributed, with the body dissatisfaction scores showing mild positive skew consistent with a non-clinical sample. Tolerance and variance inflation factor values were within acceptable limits (all VIF < 2.1), indicating no problematic multicollinearity, and inspection of residual plots supported the assumptions of linearity and homoscedasticity.
Descriptive statistics for the study variables are presented in Table 3. As noted in Chapter 3, internal consistency was good to excellent across the multi-item scales. On average, participants reported moderate levels of appearance-focused use and appearance comparison, with considerable variability, and body dissatisfaction scores spanned the full range of the measure, indicating meaningful individual differences on the outcome variable.
| Variable | M | SD | Range | α |
|---|---|---|---|---|
| Daily social media use (minutes) | 208 | 79 | 45–420 | — |
| Appearance-focused use (1–5) | 3.12 | 0.84 | 1.13–4.88 | .82 |
| Appearance comparison (1–5) | 3.28 | 0.91 | 1.00–5.00 | .94 |
| Thin-ideal internalisation (1–5) | 3.41 | 0.88 | 1.20–5.00 | .89 |
| Body dissatisfaction (BSQ) | 58.7 | 19.6 | 18–101 | .95 |
Pearson correlations among the study variables are presented in Table 4. In support of Hypothesis 1, daily social media use was positively and significantly, though weakly, associated with body dissatisfaction (r = .21, p = .004). Hypotheses 2, 3 and 4 were supported more strongly: body dissatisfaction was moderately associated with appearance-focused use (r = .46, p < .001), strongly associated with appearance comparison (r = .58, p < .001) and strongly associated with thin-ideal internalisation (r = .52, p < .001). The predictors were themselves intercorrelated, as theory would predict; in particular, appearance-focused use was moderately associated with appearance comparison (r = .49, p < .001), consistent with the proposition that appearance-focused activity provides the raw material for comparison. Notably, the correlation between daily use and body dissatisfaction was markedly smaller than those involving the quality-of-use variables, providing initial support for the central argument that the nature of use matters more than its quantity.
| Variable | 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|
| 1. Daily use | — | ||||
| 2. Appearance-focused use | .38 | — | |||
| 3. Appearance comparison | .27 | .49 | — | ||
| 4. Thin-ideal internalisation | .24 | .44 | .55 | — | |
| 5. Body dissatisfaction | .21 | .46 | .58 | .52 | — |
A hierarchical multiple regression was conducted to test Hypothesis 5 and to establish which variables uniquely predicted body dissatisfaction. At Step 1, daily social media use was entered as the sole predictor and accounted for a small but significant proportion of variance in body dissatisfaction, R² = .044, F(1, 185) = 8.55, p = .004. At Step 2, appearance-focused use, appearance comparison and thin-ideal internalisation were added to the model. The addition of these quality-of-use variables produced a substantial and significant increase in explained variance, ΔR² = .368, ΔF(3, 182) = 37.9, p < .001, with the full model accounting for 41% of the variance in body dissatisfaction, R² = .412, F(4, 182) = 31.6, p < .001.
Inspection of the standardised coefficients in the final model (Table 5) showed that appearance comparison was the strongest unique predictor (β = .38, p < .001), followed by thin-ideal internalisation (β = .24, p < .001) and appearance-focused use (β = .17, p = .012). Crucially, once these variables were entered, daily social media use no longer predicted body dissatisfaction (β = .04, p = .52). This pattern provides clear support for Hypothesis 5: the quality-of-use variables predicted body dissatisfaction well beyond the quantity of use, and the small bivariate association between daily use and body dissatisfaction was fully attenuated once appearance-focused use and the comparison and internalisation processes were taken into account.
| Predictor | B | SE B | β | p |
|---|---|---|---|---|
| Daily use (minutes) | 0.01 | 0.02 | .04 | .52 |
| Appearance-focused use | 3.96 | 1.56 | .17 | .012 |
| Appearance comparison | 8.19 | 1.48 | .38 | <.001 |
| Thin-ideal internalisation | 5.34 | 1.52 | .24 | <.001 |
In summary, all four bivariate hypotheses were supported, with the association between daily use and body dissatisfaction being weak and those involving appearance-focused use, appearance comparison and thin-ideal internalisation being moderate to strong. The regression analysis supported Hypothesis 5: the quality-of-use variables collectively explained a large and significant increment in body dissatisfaction, appearance comparison emerged as the dominant predictor, and daily use did not predict body dissatisfaction once the quality variables were controlled. The findings are interpreted in the following chapter.
The study set out to examine the relationship between social media use and body image among UK undergraduate women, distinguishing the contribution of the quantity of use from that of appearance-focused use and the associated processes of comparison and internalisation. The illustrative findings offer a coherent and theoretically meaningful picture. Consistent with Hypothesis 1, there was a positive but weak association between daily social media use and body dissatisfaction. However, in line with Hypotheses 2 to 4, appearance-focused use, appearance comparison and thin-ideal internalisation were all more substantially associated with body dissatisfaction, and in the regression model — supporting Hypothesis 5 — these quality-of-use variables explained a large proportion of variance while the quantity of use was reduced to non-significance.
The finding that appearance comparison was the strongest predictor is particularly noteworthy and dovetails with the central claim of social comparison theory (Festinger, 1954): it is not exposure to social media per se that threatens body image but the upward appearance comparisons that exposure can invite. That appearance-focused use predicted body dissatisfaction but was outperformed by comparison is consistent with the proposition that appearance-focused activity exerts its influence largely by providing opportunities for comparison. The independent contribution of thin-ideal internalisation accords with the tripartite influence model (Thompson et al., 1999) and with evidence that internalisation is a robust risk factor for body-image disturbance (Thompson and Stice, 2001). Taken together, the results support a process account of media influence in which the nature of engagement, rather than its sheer quantity, is decisive.
The attenuation of the daily-use effect once quality-of-use variables were entered is theoretically important and echoes a recurring pattern in the literature, in which global indices of use are weak and inconsistent correlates of body image while appearance-focused measures are stronger and more reliable (Holland and Tiggemann, 2016; Fardouly and Vartanian, 2016). The present illustrative data thus reinforce the argument that the common public framing — which equates time online with harm — is overly simplistic.
The relative ordering of the predictors also merits comment. That appearance comparison outperformed thin-ideal internalisation, which in turn outperformed appearance-focused use, is consistent with a chain in which appearance-focused activity supplies the opportunities for comparison, comparison activates and reinforces internalised ideals, and the resulting self-ideal discrepancy manifests as body dissatisfaction. Within such a chain, variables nearer the outcome would be expected to show the strongest direct associations, which is the pattern observed in the illustrative data. This reading coheres with the tripartite influence model’s placement of comparison and internalisation as the proximal mediators between sociocultural pressure and body dissatisfaction (Thompson et al., 1999) and with Cohen and colleagues’ (2017) finding that it is appearance-focused activity, operating through comparison, rather than overall use, that relates to body-image concern. It should be emphasised, however, that the cross-sectional data cannot establish this sequence, which is advanced as an interpretation consistent with theory rather than as a demonstrated causal pathway.
It is also worth considering why daily use retained even a weak bivariate association with body dissatisfaction before the quality variables were entered. The most plausible explanation is that heavier overall users tend, on average, to engage in more appearance-focused activity and more comparison, so that quantity of use serves as a weak proxy for the quality variables with which it is correlated. Once those quality variables were entered directly into the regression, the proxy added nothing further, and its coefficient fell to non-significance. This interpretation is consistent with the observed intercorrelations, in which daily use was modestly associated with appearance-focused use and comparison, and it explains how a variable can be a significant bivariate correlate yet a redundant predictor in a multivariate model.
The findings carry several implications for theory. First, they support the continued relevance of Festinger’s (1954) social comparison framework to the digital era and are consistent with experimental demonstrations that social media’s effect on body image is carried by comparison processes (Fardouly et al., 2015). Second, they align with the tripartite influence model’s emphasis on comparison and internalisation as the proximal mechanisms through which sociocultural pressures operate (Thompson et al., 1999), and with objectification theory’s account of how an appearance-focused, observer’s perspective on the self can sustain dissatisfaction (Fredrickson and Roberts, 1997). The pattern whereby appearance-focused use was associated with comparison, and comparison in turn was the strongest predictor of dissatisfaction, is consistent with a mediational model in which comparison transmits the influence of appearance-focused engagement — a possibility that the present cross-sectional design can suggest but not confirm.
More broadly, the results favour process-oriented over exposure-oriented accounts of media influence. They suggest that theoretical models of social media and body image should centre on what users do and how they psychologically respond, rather than on undifferentiated measures of use. This has methodological as well as theoretical implications, cautioning against reliance on global screen-time measures in future research.
The results also speak to the generalisability of theories developed largely in adolescent and non-UK samples. The pattern observed here — weak quantity effects, strong quality effects and comparison as the dominant predictor — mirrors that reported in the wider literature, suggesting that the core mechanisms specified by social comparison and sociocultural theories operate similarly among UK undergraduate women. This continuity is reassuring for the external validity of these frameworks, though it should be confirmed in more representative samples. At the same time, the prominence of peer-based comparison on social media, noted in the review, invites a modest extension of traditional sociocultural accounts, which historically emphasised professionally produced mass-media imagery: in the social media environment, the peer has become a central and perhaps uniquely potent source of appearance pressure, and theories may need to accord this source greater weight.
If, as the illustrative findings suggest, appearance comparison and thin-ideal internalisation rather than time online are the key correlates of body dissatisfaction, then interventions should target these processes. Several practical recommendations follow:
These recommendations are offered tentatively, in keeping with the correlational and illustrative nature of the study, but they illustrate how a clearer understanding of mechanisms can sharpen intervention design.
A further practical implication concerns the framing of public-health messaging. Advice that focuses exclusively on reducing screen time risks being both ineffective, if quantity is not the operative variable, and counter-productive, if it generates guilt or stigma around ordinary and often beneficial uses of social media such as maintaining friendships and accessing support. A mechanism-informed message — that the problem lies less in being online than in habitual upward comparison and the uncritical acceptance of narrow appearance ideals — is more accurate and more actionable. It also opens space for positive recommendations, such as diversifying the accounts one follows and deliberately attending to the body’s function rather than its appearance, which may be more sustainable than restriction. Platform design is a final consideration: features that quantify appearance-related approval, or that algorithmically amplify idealised appearance content, may intensify the very processes implicated here, and design choices that de-emphasise such features could plausibly reduce harm at scale, though this remains an empirical question for future evaluation.
Several limitations qualify the interpretation of the findings. First, and most fundamentally, the cross-sectional correlational design precludes causal inference. Although the results are consistent with the theorised direction of influence, it is equally possible that body dissatisfaction increases the propensity to compare and to engage in appearance-focused use, or that reciprocal and reinforcing relationships operate over time. Longitudinal and experimental designs are required to establish causality.
Second, all measures were self-report and therefore subject to common-method variance, social desirability and recall biases, particularly the retrospective estimate of daily use, which is known to correlate only modestly with objectively logged use. The appearance-focused use scale was adapted for this study and, although it demonstrated acceptable reliability, would benefit from further validation.
Third, the sample was a self-selected convenience sample of UK undergraduate women aged 18 to 25, recruited partly through social media. This raises the possibility of selection effects — for example, women more interested in or affected by the topic may have been more likely to participate — and limits generalisability to men, to other age groups, to non-students and to other cultural contexts. The study also did not assess platform-specific effects in detail or capture the content participants actually encountered.
Fourth, body image and social media use are multidimensional constructs, and the study necessarily operationalised each with a limited selection of measures. The outcome captured a single, albeit central, facet of negative body image — concern about body shape — so that related constructs such as body surveillance, appearance anxiety, drive for thinness and positive body image were not assessed. Nor did the study directly measure the potential moderators identified as theoretically important in the review, such as trait self-esteem and media literacy. A fuller model incorporating these variables, and formally testing comparison as a mediator of the link between appearance-focused use and dissatisfaction, would provide a more complete account than the parsimonious design adopted here allowed. Finally, as this is a model dissertation, the data are illustrative; the value of the work lies in demonstrating design, analysis and reporting rather than in the specific numerical estimates, and the figures should not be cited as empirical findings.
The limitations suggest clear avenues for future work. Longitudinal studies tracking appearance-focused use, comparison, internalisation and body dissatisfaction over time would clarify the direction and stability of the relationships observed here, and experimental or ecological-momentary designs could test the causal role of comparison more directly and with greater ecological validity. Incorporating objective measures of use, such as device-logged screen time and platform analytics, would address the shortcomings of self-report. Formal mediation and moderated-mediation analyses — for instance, testing whether appearance comparison mediates the link between appearance-focused use and dissatisfaction, and whether trait comparison or self-esteem moderates it — would advance the process account. Extending the research to more diverse samples, including men, gender-diverse individuals, adolescents and non-student populations, and examining platform- and content-specific effects, would improve generalisability. Finally, intervention studies evaluating comparison-focused and media-literacy programmes against screen-time-reduction approaches would test the practical implications derived from these findings.
This model dissertation examined the relationship between social media use and body image among UK undergraduate women, using a cross-sectional correlational design and illustrative data to distinguish the quantity of use from its appearance-focused quality and the associated processes of comparison and internalisation. The findings indicated that while greater overall use was weakly associated with body dissatisfaction, appearance-focused use, appearance-based social comparison and thin-ideal internalisation were considerably stronger correlates, and that the quality-of-use variables predicted body dissatisfaction well beyond the quantity of use, with appearance comparison emerging as the dominant predictor and daily use ceasing to predict dissatisfaction once the quality variables were controlled.
These results are consistent with social comparison theory, objectification theory and the tripartite influence model, and they reinforce a process-oriented understanding of how social media relates to body image: it is what young women do on social media, and how they psychologically respond, rather than how long they spend there, that is most relevant to body dissatisfaction. The practical corollary is that interventions should prioritise reshaping engagement and reducing comparison and internalisation rather than simply limiting time online. Although the study is cross-sectional and its data illustrative, it demonstrates how a theory-driven correlational dissertation can be designed, analysed and reported, and it points towards longitudinal, experimental and intervention research as the next steps in understanding and addressing this important issue for young women’s wellbeing.
British Psychological Society (2021) Code of Ethics and Conduct. Leicester: British Psychological Society.
Cash, T.F. (2004) ‘Body image: past, present, and future’, Body Image, 1(1), pp. 1–5.
Cohen, R., Newton-John, T. and Slater, A. (2017) ‘The relationship between Facebook and Instagram appearance-focused activities and body image concerns in young women’, Body Image, 23, pp. 183–187.
Fardouly, J., Diedrichs, P.C., Vartanian, L.R. and Halliwell, E. (2015) ‘Social comparisons on social media: the impact of Facebook on young women’s body image concerns and mood’, Body Image, 13, pp. 38–45.
Fardouly, J. and Vartanian, L.R. (2016) ‘Social media and body image concerns: current research and future directions’, Current Opinion in Psychology, 9, pp. 1–5.
Festinger, L. (1954) ‘A theory of social comparison processes’, Human Relations, 7(2), pp. 117–140.
Field, A. (2018) Discovering Statistics Using IBM SPSS Statistics. 5th edn. London: Sage.
Fredrickson, B.L. and Roberts, T.-A. (1997) ‘Objectification theory: toward understanding women’s lived experiences and mental health risks’, Psychology of Women Quarterly, 21(2), pp. 173–206.
Holland, G. and Tiggemann, M. (2016) ‘A systematic review of the impact of the use of social networking sites on body image and disordered eating outcomes’, Body Image, 17, pp. 100–110.
Perloff, R.M. (2014) ‘Social media effects on young women’s body image concerns: theoretical perspectives and an agenda for research’, Sex Roles, 71(11–12), pp. 363–377.
Thompson, J.K., Heinberg, L.J., Altabe, M. and Tantleff-Dunn, S. (1999) Exacting Beauty: Theory, Assessment, and Treatment of Body Image Disturbance. Washington, DC: American Psychological Association.
Thompson, J.K. and 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(5), pp. 181–183.
Tiggemann, M. and Slater, A. (2013) ‘NetGirls: the Internet, Facebook, and body image concern in adolescent girls’, International Journal of Eating Disorders, 46(6), pp. 630–633.
Tiggemann, M. and Zaccardo, M. (2015) ‘”Exercise to be fit, not skinny”: the effect of fitspiration imagery on women’s body image’, Body Image, 15, pp. 61–67.
You May Also Like
All work is written by human writers. 100% AI free, guaranteed.
100% money back guarantee if you find plagiarism in our work.
COMPANY DETAILS
