October 26, 2026
Instagram algorithms amplify commercial exposure and consumption stimuli
Alcides Gomes Beato Neto; Heinrich Da Solidade Santos
DOI: 10.22167/2675-6528-202600485
Article derived from a Course Conclusion Work (TCC), with content based on the student’s original work and adapted to the editorial format of the E&S Magazine with the support of the ResumeAI tool, an artificial intelligence solution developed by Instituto Pecege for textual synthesis and organization.
Summary
Instagram has consolidated itself as a platform that brought together social interaction, personalized recommendation, advertising, and shopping features in a single usage flow, becoming a relevant case for discussing digital addiction and consumerism from a software engineering perspective. The objective was to analyze how infinite scrolling, recommendation algorithms, and the platform’s commercial stimuli influenced user engagement, the presence of commercial content, and the stimulation of consumption in the analyzed environment. The research was classified as applied, exploratory, and descriptive, with a mixed approach and experimental design. A literature review, documentary analysis of official materials from Instagram and Meta, and a computational simulation with synthetic data were conducted, in which classification parameters were manipulated and their effects on content ranking scenarios were observed. The results indicated that the combination of interface continuity, algorithmic personalization, and the insertion of commercial signals significantly increased the presence of content with the potential to stimulate purchasing and social comparison, while session depth showed more moderate variations across scenarios. It was concluded that the problem was linked not only to individual user behavior but also to design decisions that prioritized retention and commercial value over digital well-being criteria.
Keywords: Digital well-being; Social comparison; Persuasive design; Content personalization; Infinite scroll.
1. Introduction
Instagram has consolidated itself as a platform that transcends its original function of photo sharing. Over time, the app has evolved into a multifaceted environment, integrating personal image production, entertainment, content recommendation, advertising, and shopping features into a single usage flow. This transformation has established it not only as a space for socialization but also as a showcase of lifestyles, an influence channel, and a consumption environment. Given this complexity, Instagram has become a relevant object of analysis for understanding how software engineering decisions can influence user behavior, attention, and consumption choices.
Recent literature has deepened the discussion on the effects of intensive social media use on well-being. Systematic reviews and meta-analyses have indicated that problematic use of these platforms is associated with worse mental health indicators, especially among adolescents and young adults (Faelens et al., 2021; Shannon et al., 2022). In the context of Instagram, this problem is intensified by the image-based usage experience, which fosters social comparison, displays visible popularity metrics, and offers constant reinforcement through likes, comments, and recommendations. Thus, the concern is not limited to the time a person spends on the app, but also to how the platform organizes this time and the stimuli that arise during navigation.
Beyond the behavioral dimension, Instagram operates with a prominent economic dimension, where digital influence, targeted advertising, and the desire for consumption are deeply integrated. Recent studies demonstrate that more intense use of social media may be related to an increase in materialistic values, the desire to keep up with idealized lifestyles, and a greater propensity for impulsive buying, especially when content creators and digital influencers actively participate in this process (Djafarova; Bowes, 2021; Yang et al., 2024; Ozimek et al., 2024; Dinh; Lee, 2024; Shamim et al., 2024). This dynamic positions Instagram not just as a social network, but as a digital system that connects user attention, desire, and commercial conversion.
From a software engineering perspective, this discussion is particularly relevant because user behavior is not merely a matter of individual will, but is intrinsically influenced by choices in product architecture, interface design, and algorithmic logic. The platform’s official documentation, for example, reveals that different sections, such as Feed, Stories, Explore, and Reels, utilize distinct signals to organize the displayed content. Content recommendation relies on machine learning models trained to predict future actions and order items based on relevance, probability of interaction, and other indicators (Instagram, 2023; Instagram, 2025; Vorotilov; Shugaepov, 2023; Levis et al., 2025). When this logic is combined with infinite scrolling, the user experience tends to eliminate natural stopping points, encouraging continuous consumption and increasing repeated exposure to inspirational, commercial, and highly engaging content (Rixen et al., 2023). This interconnection between design and behavior reveals an important gap in understanding how technical decisions impact well-being and digital consumerism.
The relevance of this study lies in the need to understand the implications of software design choices that prioritize retention and commercial value, often at the expense of digital well-being criteria. The problem is not restricted to individual behavior but extends to design decisions that shape the user experience. Considering this scenario, this work aimed to analyze how the combination of infinite scrolling, recommendation algorithms, and commercial stimuli on Instagram influenced user engagement, increased commercial exposure, and reinforced consumption stimuli, based on an experimental study with a literature review, documentary analysis, and simplified algorithmic simulation.
2. Material and Methods
The research was classified as applied, as it discussed a real digital product problem with technical, economic, and social implications. Regarding the objectives, the study was characterized as exploratory and descriptive, seeking to understand the relationship between infinite scrolling, algorithmic recommendation, excessive use, and consumerism on Instagram, in addition to describing the platform’s functioning. A mixed approach and experimental design were adopted, with the manipulation of parameters in simulated scenarios and the observation of effects on the variables of interest.
The construction of the theoretical base was carried out through a literature review of scientific articles. The topics covered included problematic use of social media, persuasive design, social comparison, digital consumerism, and recommender systems. Complementarily, a documentary analysis of institutional materials from Instagram and Meta was conducted, focusing on texts about content ranking, recommendation, and terms of use. This stage allowed for the relationship between academic literature and the information disclosed by the platform itself about its functioning.
For the computational simulation, synthetic data were generated and routines were executed in Python. Custom computational routines were used for pseudo-random sorting, organization of experimental scenarios, and calculation of item scores. This approach was chosen to allow controlled repetition of procedures, manipulation of parameters, and comparison between different combinations of weights and variables, ensuring the standardization of the experiment. Reproducibility was ensured by the use of a fixed seed for pseudo-random generation.
The treatment of simulated data and the consolidation of results were carried out with the support of the NumPy and pandas libraries. NumPy was employed in the generation of beta distributions and in the numerical operations of the simulation. Pandas, in turn, was used in the organization of simulated sessions, in the calculation of aggregated metrics, and in the export of final tables in tabular format. This computational environment contributed to the organization and reproducibility of the experiment.
The values of the variables were produced by beta distributions, selected for their suitability on a limited scale between zero and one and for their ability to generate asymmetries compatible with different content and user profiles. The variables included novelty, which represented the item’s potential for surprise to the user; emotional intensity, referring to the item’s capacity to generate affective arousal; social proof, which indicated the level of perceived popularity of the content; and commercial appeal, related to the presence of brands, products, or purchase triggers.
Other variables considered were relational proximity, which measured the degree of social connection to the content’s origin; purchase affinity, which denoted the individual’s predisposition to respond to commercial content; social comparison, which reflected the user’s sensitivity to inspirational signals and social metrics; and self-control, which represented the ability to interrupt the session. These variables were defined to influence item scoring and the probability of session interruption.
The simulation involved 1,000 independent sessions, each with up to 15 consumption blocks and 20 candidates per block. The final score for each item resulted from the weighted combination of relational proximity, novelty, emotional intensity, social proof, and commercial appeal adjusted by purchase affinity. In each block, the 20 candidates were compared by score, and the item with the highest score was considered consumed in that cycle.
Four scenarios were defined for comparative analysis. The reference scenario prioritized novelty and relational proximity. The engagement scenario expanded the weights of variables linked to the user’s expected response. In the commercial signal scenario, greater weight was given to commercial appeal and content with purchasing potential. Finally, the friction mitigation scenario maintained the logic of the commercial signal scenario but introduced artificial friction points every five blocks consumed.
The session continuity was updated with each consumed block, based on a logistic function that combined the score of the selected item, the user’s social comparison, block progression, and self-control. The probability of interruption was recalculated at the end of each block, considering self-control and the accumulated attractiveness of the feed, which allowed estimating the average session depth in each scenario. As the study did not involve human beings, interviews, questionnaires, or personal data, there was no need for submission to the Ethics Committee.
3. Results and Discussion
The results of this research revealed that Instagram’s software architecture, by integrating infinite scrolling, recommendation algorithms, and commercial stimuli, significantly influences the user experience, especially regarding exposure to content with potential for consumption and social comparison. Document analysis, literature review, and computational simulation converged to indicate that the problem of digital addiction and consumerism is not restricted to individual behavior, but is intrinsically linked to design decisions that prioritize retention and commercial value. This interconnection between design and behavior was the central focus of the investigation, seeking to unravel how the platform organizes the content flow and the stimuli that arise during navigation.
Instagram as a retention architecture
The documentary analysis of official materials from Instagram and Meta demonstrated that the platform is conceived as a highly personalized service, where content, recommendations, and advertisements are carefully selected for each user. This selection is based on a wide range of activity and relevance signals, as described by the company itself (Instagram, 2025). This approach underscores that the discussion about user retention and consumption stimulation cannot be isolated from the platform’s design, as the navigation environment is actively organized to optimize the individualized experience.
The examined materials also showed that the different sections of the application, such as Feed, Stories, Explore, and Reels, operate with distinct classification logics. Although all are fed by similar behavioral signals, each surface has its own mechanism for ranking content. This diversity of mechanisms contributes to the fluidity and adaptability of the user experience, making the application seem constantly adjusted to the patterns and interests of each person, which is a key factor for retention.
In the specific case of Explore, Meta’s engineering detailed a multi-stage funnel, which includes candidate retrieval, initial ranking, subsequent ranking, and final re-ranking, using machine learning models to predict future actions (Vorotilov; Shugaepov, 2023; Levis et al., 2025). This algorithmic complexity is crucial for understanding how the platform manages to hold user attention for long periods. With each user action, the system seeks a dual function: to express a present behavior and, at the same time, provide data to organize future behaviors, creating a reinforcement cycle that amplifies similar interests and profiles.
Infinite scrolling, a mechanism for frictionless continuity, was identified as a central element in this retention architecture. In the human-computer interaction literature, infinite scrolling is recognized for eliminating natural stopping points, which favors continuous consumption and increases repeated exposure to inspirational, commercial, and highly engaging content (Rixen et al., 2023). This design feature, by removing clear stopping landmarks, causes the content flow to be perceived as a single sequence, facilitating the prolongation of navigation without conscious user effort.
Monge Roffarello and De Russis (2023) classified infinite scrolling as an attention-capturing pattern that reduces the user’s temporal autonomy. The user experience, therefore, is not limited to the amount of content available, but to how the interface guides navigation, making natural pauses and termination decisions difficult. The effect is intensified because scrolling does not require deliberate action, being perceived as a continuous flow of rewards. This combination of infinite scrolling with an adaptive sequence of content makes the experience immersive and often difficult to interrupt.
This interface choice, although seemingly neutral, has broad behavioral consequences, such as prolonged engagement and use without clear intent. Fagan (2024) describes this phenomenon as a “gray zone” between legitimate and dark persuasion, where features like choice defaults and visual seductions become problematic when they are not transparent and difficult to avoid. On Instagram, convenience, personalization, and commercial incentives combine in an experience that, by design, seeks to maximize retention.
From personalization to reinforcement: the algorithmic loop of engagement
Meta’s technical literature and official materials allow for describing Instagram’s functioning as a four-step algorithmic loop. Firstly, the system collects user events, such as time spent viewing, likes, comments, saves, shares, profiles visited, and searches. Next, this information feeds models that estimate the probability of new interactions. In the subsequent step, content is ranked according to the objective function chosen by the platform. Finally, the behavior generated by the classification itself returns to the system as a new learning signal, feeding back into the cycle.
This loop, while useful for reducing informational overload and finding relevant content, becomes problematic when the system’s dominant objective is retention, session depth, or commercial value. In these cases, ranking tends to favor content that maximizes immediate response, rather than necessarily that which would be healthier, more diverse, or balanced for the user. The prioritization of engagement metrics can thus lead to a bias in content selection.
In visual social media, such as Instagram, this issue intensifies because many engagement signals are linked to social comparison and self-assessment. Faelens et al. (2021) highlighted that a large part of the literature on Instagram use addresses social comparison, body image, and psychological well-being. This means that the same system that learns to capture the user’s attention can also inadvertently reinforce content that activates known vulnerabilities, such as dissatisfaction with one’s own image or the fear of missing out (FOMO).
Infinite scrolling, for example, has the predominant technical objective of reducing navigation friction and increasing session continuity. The probable effect on the user is a lower perception of closure and more automatic navigation, with the associated risk of prolonged stay and use without clear intention. Personalized ranking, in turn, aims to order items with a higher probability of positive response, resulting in a content sequence more aligned with individual history, but with the risk of reinforcing preferences and reducing content diversity.
Public popularity metrics, such as likes and comments, aim to increase social signal and visible feedback, which can lead to social comparison and validation by numbers. The associated risk includes anxiety, fear of missing out (FOMO), and intense self-assessment. The integration of content creators with brands seeks to increase commercial conversion without disrupting the flow, promoting a perception of naturalness in consumption recommendations, but with the risk of less distinction between social content and advertising.
Finally, personalized ads and offers aim to maximize commercial relevance and advertising return. This results in recurrent exposure to products compatible with the user’s inferred interest, with the risks of impulse buying and consumerist saturation. This articulation between interface design, objective function, and monetization logic demonstrates how the final user experience is shaped by software engineering choices.
Instagram and the rise of consumerism
Instagram stimulates the desire to buy not only through explicit advertisements but also through indirect channels, according to the literature. The platform values polished images, inspirational lifestyles, the logic of followers, and the constant presence of influencers and brands. In this environment, consumption is presented not just as a transaction, but as part of the construction of personal identity, which makes it a more subtle and integrated persuasive mechanism into the user experience.
Djafarova and Bowes (2021) demonstrated that Instagram marketing strategies influence impulse purchases among Generation Z consumers. They highlighted the strength of visual content, familiarity with influencers, and the appearance of spontaneity as crucial factors. Shamim et al. (2024) corroborated this, observing that the credibility of influencers and sponsored posts increased purchase impulse. These findings indicate that commercial conversion on Instagram transcends traditional advertising, being driven by a more complex and integrated dynamic.
This discussion is supported by broader studies on materialism and social media. Yang et al. (2024) identified an association between social media use and materialistic values, with self-control and self-acceptance playing important roles. Ozimek et al. (2024) found that a greater materialistic orientation relates to more symptoms of stress and social media addiction. Together, these studies suggest that the digital environment can reinforce both the desire for consumption and the subjective weariness associated with this process.
Okazaki et al. (2021) added that, in the context of Instagram, engagement with brands can evolve into compulsive use of the network and, subsequently, into compulsive buying, mediated by “technostress”. This result is relevant for dispelling the idea that the problem is limited to the effectiveness of advertising. Instead, it is a broader dynamic, where entertainment, brand, affect, and habit mix, creating a reinforcement cycle that is difficult to break.
From a software perspective, Instagram uses at least four mechanisms that favor this process. The first is relevance personalization, which increases the compatibility between the displayed content and the user’s probable interest. The second is social proof, which legitimizes desire through likes, views, comments, and the authority of content creators. The third is interface fluidity, which reduces reflection time between one stimulus and another. The fourth is the integration between organic content, advertising, and “social commerce”, which diminishes the symbolic separation between seeing, desiring, and buying.
Documentary analysis reinforces this interpretation, as Instagram’s Terms of Use (Instagram, 2025) explicitly state that advertisements, offers, and other sponsored content are an integral part of the service and are personalized for each person. This means that monetization is not an external layer, but is integrated into the normal functioning of the platform. Therefore, the discussion about consumerism on Instagram is fundamentally a discussion about structural product and software engineering decisions.
Simplified algorithmic model of the case
Based on the identified elements, a simplified model was developed to represent the case, preserving three central characteristics of Instagram. The first is the existence of a large set of candidates available for display. The second is the calculation of engagement predictions based on user and item signals. The third is the possibility of shifting the ranking according to the weight assigned to the commercial signal. The objective was not to reproduce the real system, but to create a didactic instrument to observe how changes in objectives alter the composition of the feed.
For the simulation, 1,000 independent sessions were performed, each with up to fifteen consumption blocks and twenty candidates per block, totaling a sufficient data volume to compare scenarios without compromising the model’s simplicity. Item scoring was built by a weighted combination of simulated variables: novelty, emotional intensity, social proof, relational proximity, and commercial appeal adjusted by purchase affinity. In each block, the twenty candidates were compared by score, and the item with the highest score was considered consumed in that cycle.
In the reference scenario, the simulation assigned greater weight to novelty and relational proximity. In the engagement scenario, the weights of variables related to the user’s expected response were amplified. In the commercial signal scenario, a higher weight was added to commercial appeal and content with greater purchase potential. This structure allowed for a clear comparison of results. The probability of interruption was recalculated at the end of each block, combining the simulated user’s self-control with the accumulated attractiveness of the feed, which allowed for estimating the average session depth in each scenario.
The main variables used in the simplified algorithmic model included novelty, which represents the item’s potential for surprise to the user and increases reward expectation. Emotional intensity, which measures the item’s ability to generate affective arousal, elevates the probability of retention. Social proof, which indicates the perceived popularity level of the content, increases social validation and curiosity. Commercial appeal, which denotes the presence of brands, products, or purchase triggers, elevates the item’s commercial value.
Relational proximity, which is the degree of social bonding with the content’s origin, favors relational interest. Purchase affinity, which represents the individual’s predisposition to respond to commercial content, interacts with commercial appeal. Finally, social comparison, which measures user sensitivity to inspirational cues and social metrics, reduces the probability of interruption. All these variables were defined in a range from zero to one, allowing for controlled manipulation of parameters in simulated scenarios.
Simulation results
The simulation results revealed significant differences between the scenarios, mainly in feed composition and consumption stimulus. In the reference scenario, which prioritized relational and temporal aspects, the average session depth was 4.12 blocks, with an estimated time of 2.1 minutes. The average commercial content participation in this scenario was 15.5%, and the average purchase stimulus index was 0.460. The final average interruption probability was 0.247.
In the engagement-optimized scenario, the average session depth increased slightly to 4.29 blocks, maintaining the estimated time at 2.1 minutes. However, the presence of commercial content increased expressively to 47.9%, and the purchase stimulus index to 0.568. The final average probability of interruption remained close, at 0.246. This result suggests that engagement optimization altered the quality of the displayed content more sharply than the session duration.
In the scenario optimized by engagement with commercial signal reinforcement, the average session depth dropped to 3.54 blocks, with an estimated time of 1.8 minutes. However, commercial content participation reached 94.2%, and the purchase stimulus index increased to 0.695. The final average probability of interruption rose to 0.299. This data indicates that prioritizing commercial signals drastically intensifies exposure to consumption content, even if it does not result in longer sessions.
This pattern is crucial because it demonstrates that the main effect of algorithmic adjustments, in this experiment, was not a substantial increase in session duration, but rather a relevant change in exposure quality. The scenarios more oriented by engagement and commercial signal altered the type of content displayed more intensely than the total user dwell time. This suggests that, in systems like Instagram, commercial value can be amplified not only by prolonging usage, but by reorganizing the content flow towards stimuli more compatible with consumption and social comparison.
Critical discussion and implications for software engineering
The discussion of the results confirms that Instagram’s problem cannot be attributed to a single isolated component. The platform becomes more prone to inducing problematic use when three elements act together: an almost frictionless continuity interface, algorithms trained to predict and reinforce engagement, and a monetization integrated into the navigation flow itself. It is this combination that amplifies commercial exposure, reinforces aspirational stimuli, and increases consumption pressure, even if the effects on retention may vary among the simulated scenarios.
For software engineering, this analysis brings at least five important implications. The first is the need to revise success metrics. Indicators such as time spent, session depth, and return rate are useful for business, but are insufficient as the sole criterion for quality. A system can retain the user for a long time and, at the same time, produce unwanted effects on autonomy, well-being, and consumption patterns, requiring a more comprehensive view of the product’s impact.
The second implication concerns product transparency. Although Meta already offers public descriptions of Instagram’s general functioning, which represents an advance compared to complete opacity, these descriptions remain at a high level. They do not concretely clarify how conflicts between retention, content diversity, commercial exposure, and user well-being are resolved. For products with significant social impact, this limit of transparency is a relevant point to be improved, allowing for a deeper understanding of design choices.
The third implication is that friction mechanisms, although technically simple, can be important. Human-computer interaction studies, such as those by Monge Roffarello and De Russis (2023), show that small contextual interruptions can reduce automatic navigation and give the user back some sense of control. In the context of this work, this suggests that pause reminders, stopping points, or continuity confirmations can function as complementary measures of a more responsible design, mitigating excessive continuity.
The fourth implication is that recommendation models should incorporate diversity and well-being as explicit reclassification constraints. If the system learns that inspirational and consumptive content generate more response, the natural tendency is to reinforce them. A more responsible reordering could limit the excessive concentration of certain stimuli, better distribute content types, and reduce harmful reinforcement cycles, promoting a more balanced experience that is less prone to vulnerabilities.
The fifth and broadest implication is that software engineering cannot treat digital addiction as an inevitable externality. When a product continuously measures, tests, and adjusts its experience to maximize retention, the effects of problematic use cease to be a purely individual accident. They become part of the predictable risks of the system’s architecture itself, requiring developers to consider the ethical and social implications of their design choices from the initial stages of the project.
Complementary scenario of friction mitigation
As a complementary exercise, a fourth scenario was executed in the simulation, in which the ranking logic with commercial signal was maintained, but artificial friction points were inserted every five blocks consumed. These points represented pause reminders, changes in feed rhythm, or the need for confirmation to continue navigation. The intention was to observe if small interface barriers could reduce the automatic continuity of the session and, consequently, the user’s permanence on the platform.
In this friction mitigation scenario, the average session depth dropped to 3.40 blocks, and the average final interruption probability increased to 0.311. The estimated session time was 1.7 minutes. Although friction reduced dwell time when compared to other scenarios, the proportion of commercial content remained very high, reaching 94.3%, and the purchase stimulus index was 0.696. This data indicates that the interface layer, by itself, does not sufficiently alter the overall exposure logic when the classification continues to direct the feed towards consumerist stimuli.
This finding reinforces an important conclusion for software engineering: interventions solely on the interface help, but do not replace changes in the algorithmic layer. When the recommendation architecture strongly favors consumptive signals, pause reminders act as partial buffers, but do not resolve the root of the problem. For a more effective change and a significant impact on digital well-being, it would be necessary to revise the system’s objective itself and the weights assigned to content classification, shifting the focus from purely commercial metrics to include criteria of quality and diversity.
In summary, the research demonstrated that the combination of infinite scrolling, algorithmic personalization, and commercial stimuli on Instagram resulted in an expressive increase in content with the potential to stimulate purchasing and social comparison. Although session depth showed more moderate variations between scenarios, the quality of exposure was drastically altered towards a more consumerist environment. This reinforces the understanding that the problem is linked not only to individual user behavior but also to design decisions that prioritize retention and commercial value above digital well-being criteria, responding directly to the study’s objective.
4. Conclusion
This study aimed to analyze how infinite scrolling, recommendation algorithms, and Instagram’s commercial stimuli influenced user retention, the presence of commercial content, and the stimulation of consumption on the platform. It was found that the combination of interface continuity, algorithmic personalization, and the insertion of commercial signals significantly increased the presence of content with the potential to stimulate purchases and social comparison. Simulation results indicated that, although session depth showed more moderate variations across scenarios, exposure quality was drastically altered towards a more consumerist environment. This dynamic highlights that the problem is not restricted to individual user behavior but is intrinsically linked to software engineering design decisions that prioritize retention and commercial value over digital well-being criteria. The main contribution lies in demonstrating, in an applied manner, how product architecture and interface design choices shape the user experience, directly influencing consumerism and digital well-being.
Despite the relevant findings, the study presented limitations, such as the use of a synthetic algorithmic model that did not faithfully reproduce Instagram’s proprietary system and the focus on a single platform, which does not exhaust the discussion on visual social networks. For software engineering, the results imply the need to revise success metrics, explicitly incorporating digital well-being, transparency, and content diversity as quality criteria. It is suggested that friction mechanisms, such as pause reminders, can act as complementary measures, but they do not replace the revision of system objectives and the weights assigned to algorithmic ranking. Future studies could expand the experiment with new ranking scenarios, test other combinations of weights and variables, and incorporate additional friction mechanisms in the interface. It is also recommended to develop research with observational data, public databases, or user participation, when methodologically feasible and ethically approved, to verify the proximity of the simulated patterns to the real behavior of navigation and consumption on social networks.
Bibliographic References
Djafarova, E.; Bowes, T. 2021. Instagram made me buy it: Generation Z impulse purchases in fashion industry. Journal of Retailing and Consumer Services 59: 102345.
Faelens, L.; Hoorelbeke, K.; Cambier, R.; Van Put, J.; Van de Putte, E.; De Raedt, R.; Koster, E.H.W. 2021. The relationship between Instagram use and indicators of mental health: a systematic review. Computers in Human Behavior Reports 4: 100121.
Ozimek, P.; Brailovskaia, J.; Bierhoff, H.-W.;
Shannon, H.; Bush, K.; Villeneuve, P.J.; Hellemans, K.G.C.; Guimond, S. 2022. Problematic social media use in adolescents and young adults: systematic review and meta-analysis. JMIR Mental Health 9(4): e33450.
Yang, Q.; Xu, Y.; Van den Bos, K. 2024. Social network site use and materialistic values: the roles of self-control and self-acceptance. BMC Psychology 12(1): 55.
Article originating from the Final Course Work of the Specialization in Software Engineering of the MBA USP/Esalq
To learn more about the course, click here and access the MBX Academy platform