A scoping review of current research on social media use among children and adolescents

Social media has changed how people interact. With increased access to advanced Information and Communication Technologies (ICT’s), concerns with possible impacts on health, learning and interpersonal social interaction has grown. While the emergence and increased use of ICT’s and social media has been noted within all age groups, use and impact of social media has been especially debated in relation to the lives of young users. The generation of children growing up today is unique in terms of the technologies they have access to and use. In 2023, almost all children in the ages between three and seventeen had been online, with some of the youngest, five- to seven-year-olds, increasing their online presence over the last few years [1]. Historically, tablets have been the main devices used by the younger children engaging in online activities, while young users of ages between twelve and seventeen have been, and still are, more likely to use mobile phones when going online (ibid). By the age of eleven, nine out of ten children of today own their own mobile phone (ibid.). Despite the minimum age for creating an account being thirteen on every major social media platform (TikTok, Instagram, Twitter, Pinterest, Kik, YouTube, Snapchat and Facebook), studies have shown that a great majority of users of these platforms are younger and that numerous minors find ways to create accounts for themselves even when they have not reached the required age [2]. Despite the age requirement of thirteen, 50% of children aged  three to twelve uses at least one social media app or visit a social media site regularly (ibid.).

Although children and adolescents use ICT’s and social media heavily, the body of work on how they spend time with these technologies, how they make meaning of them, and how ICT use affect their social lives, learning practices et cetera is still growing and findings are ambiguous and sometimes even contradictory. While the possible effects of screen time and use of ICT’s in general and social media in particular might have on children and adolescents is publicly debated, the empirical evidence supporting either of the often-polarized perspectives that are given space in media, is limited. The critical voices arguing for limiting children’s screen time based on a concern about its possible impact on children’s development, social life and well-being often argue referring to studies of limited scientific quality where causal relationships between screen time and impaired health is missing. For example, it has previously been highlighted through meta-analyses that existing work on ‘screen time’ have been lacking in differentiating various forms of interaction and technologies, and the fact that different mediated activities entail different consequences, has been neglected [3]. Also, it has been proven to be difficult to isolate the use of ICT’s as being a single impacting factor and show precisely how use of these technologies alone affects young people’s well-being. The negative effects that have been demonstrated are also very limited and it has been difficult to investigate the causality between different factors involved [3]. While meta-analyses have been criticized for possible overreliance on bivariate correlations rather than controlled standardized regression coefficients[4], recent meta-analysis studies of social media use have shed light on the relationship between screen time and well-being and have concluded that there is no evidence supporting the claim that time spent on social media is correlated with adolescent mental health problems [4, 5], nor that associations between adolescents’ technology engagement and mental health have increased over time [6]. Since causal relationship are yet to be demonstrated, it may be that extensive screen time leads to a decline in experienced social and mental well-being, but it may also be that an already experienced decline in social and mental well-being may lead to watching more TV, playing more videogames, or increasing the time one spends in social media.

The complexity of these issues is further compounded by the emotions and perceptions of parents regarding their children’s screen time. While users of all ages tend to describe conflicting emotions related to their screen time [7], parents specifically often describe feeling shame for allowing their children ‘too much’ screen time. They often struggle to define what ‘too much’ really means or to understand where this feeling of shame originates [8]. According to often referenced guidelines developed by WHO, the time children, especially the youngest ones, spend with screens should be very limited [9]. However, the main message from of these guidelines, which regard physical activity, sedentary behavior and sleep for children under five years of age, is that it is time spent being physically inactive that should be limited, regardless of whether that happens in front of a screen or not. In addition to these guidelines often being misinterpreted, critique has been raised against them, and similar guidelines, emphasizing the need for guidelines that are more specific as well as more clearly evidence-based. For instance, Przybylski and Weinstein [10] has questioned the evidence base for strict screen time limits and suggested that screen use is not inherently harmful and argued for more context-specific guidelines. Also, LeBlanc et al. [11] suggested that cultural and environmental factors significantly influence screen time and that guidelines therefore should be more adaptive to these variances.

For this scoping review, the objective is to contribute to the ongoing discussion on social media use among children and adolescents by providing an overview of conduct, origin and distribution among existing peer-reviewed research on this very topic. This paper focuses on a specific form of screen time, namely social media use, by asking: What are the key themes in terms of conduct, findings and gaps in previous peer-reviewed research on children’s and adolescents’ social media use?

Scoping reviews have been described as “a type of knowledge synthesis that uses a systematic and iterative approach to identify and synthesize an existing or emerging body of literature on a given topic” [12, p. 565]. Common reasons for conducting scoping reviews are wanting to examine the extent, range, and nature of research activity, determining the value of undertaking a full systematic review, summarizing, and disseminating research findings and identifying research gaps in existing literature [13, p. 21]. This review was conducted systematically following the steps involved in the established and commonly applied framework provided by Arksey and O’Malley [13], i.e. identifying the research question, identifying relevant studies, selecting studies, charting the data and collating, summarizing and reporting the results [13]. Additionally, the Preferred Reporting Extension for Systematic Reviews and Meta-Analyses (PRISMA) [14] has been used to ensure the providing of a high degree of systematic conduct and methodological detail in the documentation of the ‘audit trail’ [15], in this case referring to the process of identifying, screening and including studies for this review.

For this scoping review, the aim is to provide an overview of the nature of current peer-reviewed research on social media use among children and adolescents, making a scoping review endeavor a suitable approach. As the aim and research question had been identified, the next step following the scoping review framework was to identify relevant studies and select studies to be included in the review. Which studies to include was decided through formulating and following a set of inclusion and exclusion criteria. Inclusion criteria for articles to be eligible to be included in the review were: that the study involved social media use among children and adolescents (0–18 years of age), that the article was peer-reviewed and published in a scientific journal or academic conference proceedings, and that it presented empirically based results. One of the first criterions formulated was that “social media use” had to be mentioned within the abstract along with at least one of the search terms used for identifying social media use among this specific use group of children and adolescents. Studies were required to be written in English. Due to the specific interest in peer-reviewed work, grey literature was not included. Also, theoretical papers and opinion and perspective pieces not involving any empirical data were not included in the review. Yet, review studies were.

While the focus of this review is on use of social media among children and adolescents, it is not limited to only included the perspectives of the users themselves, but also to allow other perspectives on their use. Therefore, the criteria for study population were not limited to involve only children and adolescents themselves, but also to include other stakeholders, such as parents and other caregivers, teachers, healthcare professionals and the similar who contribute with additional perspectives on children’s and adolescent’s use of social media. However, while the perspectives of these populations added additional insights to a broader understanding, the focus of this review is to map out what is known about social media use among children and adolescents, not among parents and other caregivers, teachers or healthcare professionals. Therefore, studies focusing on use of social media among these external stakeholder user groups were not included. For instance, a study involving parents reasoning on possible impact on family functioning by social media use by depressed adolescents was included [16], yet another study exploring how parents of children with special needs seek out Facebook groups for social support and help in performing caregiving work [17] was not. Another example of an article that met this inclusion criteria was one based on a study exploring adults’ perceptions of how sexualized images on social media might influence adolescent girls’ mental health by conducting in-depth interviews with parents, school support service staff and youth mental health service providers [18]. Conversely, another study on social media use among parents and caregivers of children with cancer [19] did not meet the inclusion criteria and was thereby excluded.

The literature search was conducted during early March 2024. The electronic databases Scopus, Web of Science and ACM Digital Library were searched with support of a university librarian to identify relevant studies. This initial search resulted in a total of 1161 identified articles. The search focused on use of social media and youth, using variations of the search string (minor* OR child* OR kid*) AND "social media use"’. In one of the earliest conducted searches, “social media” was used as a search term, yet was quickly identified as being too broad, why the search term was narrowed to involve ‘social media use’ specifically. All peer-reviewed records, published at any point in time, where both ‘social media use’ and one of the youth search terms were identified in the abstract were included. Mendeley was used for management of the bibliographic data and the web application tool Rayyan was used to support screening. However, this tool was only used to organize and support manual work, not for automatic screening.

For the final step in the Arksey and O’Malley [13] framework, i.e. charting the data and collating, summarizing and reporting the results, an Excel data-charting form was designed for extracting data. This was done prior to reading the final full text articles included in the review. Initial general extraction categories were created influenced by common categories identified by Mak and Thomas [12] and involved: authors, year, geographical location, study population, main results, study limitations and future directions (ibid.). In addition, and as suggested by the same authors, additional extraction categories were designed to capture data extraction relevant for the specific research question (ibid.). These categories included: academic discipline, source type, author keywords, methodological approach, data collection design, research question and aim, and age of children and adolescents involved. Also, additional subcategories were iteratively added to identify self-report studies, longitudinal studies and cross-sectional studies, to specify both geographical location of first author and of where the study was conducted, and whether the study population representatives had been identified with any medical disorder, syndrome, or special needs. Also, a column for adding additional comments and reflections was included.

In comparison to when conducting a systematic review, the aim of a scoping review is not to seek to synthesize evidence or aggregate findings from different studies, but rather to present an overview of all material reviewed [13]. Hence, a common approach when analyzing and presenting results from scoping reviews is to follow a dual approach involving both a numerical and thematic analysis [12, 13]. Thus, as the data had been extracted from the 244 articles eventually included in this review, both numerical and thematic analyses were conducted. The numerical analysis involved mapping and visualizing the quantifiable results (such as number of studies conducted per year, country of origin, discipline of origin and methodological approach) in tables, figures and charts. The thematic analysis was conducted using the framework developed by Braun and Clarke [20]. This involved iteratively developing codes to categorize different data segments relevant to the review’s aim and scope, grouping these codes, and developing themes that represent findings or ‘patterns across the dataset’ [12, p. 566]. This part of the analysis was, and usually is, very time-consuming due to the complex material and varied studies involved, making comparisons in-between them sometimes challenging [13, 21]. NVivo was used for coding and for running word frequency queries. The word cloud in ‘Fig. 4. Visualization of author keywords’ was generated using the Microsoft Word plugin Bjorn’s Word Clouds. Language editing has been assisted by ChatGPT (provided by OpenAI).

Figure 1 represents the PRISMA flowchart for the study selection process and provides details on the identified, screened, included, and reviewed articles for this scoping review.

PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flowchart for the studies identified, screened, included, and reviewed

Full size image

The literature was searched within three electronic databases and resulted in a total of 1161 initial articles. Out of those 1161 articles, 203 were upon downloading automatically identified by Mendeley as duplicates and were manually controlled and removed by author (not by software) before the screening of abstracts begun. 197 were removed by author after a screening of titles (mainly removing results of using ‘minor*’ as a search term, resulting in articles focusing on ‘minority’ and ‘minorities’, which falls outside of the scope of this review), leaving 761 records to be exported to Rayyan for screening. During this next step of the process, the abstracts of those 761 records were screened and tentatively coded by the author, leaving 440 articles to be sought for retrieval (with three records not possible to retrieve). 437 references were exported from Rayyan back to Mendeley, where the full text articles were downloaded to be read and assessed for relevance and eligibility. Out of those 437, 188 were excluded (reasons were: wrong population (n = 86), wrong focus (n = 52), wrong publication type (n = 36), wrong unit of analysis (n = 15), background article (n = 2) and foreign language (n = 2)) leaving a remaining 244 articles to be included in the review.

The review shows that there has been a steady increase in articles on social media use among children and adolescents over the last years with most of the studies having been conducted during the last five years, possibly being a reflection of the rise in social media use during the last decade (see Fig. 2).

Number of publications across time

Full size image

Geographically, the research encompassed in this scoping review has predominantly been conducted in Western nations, with the majority of articles emanating from the United States (33.2%, n = 81), followed by the United Kingdom (9.8%, n = 24), Australia (8.2%, n = 20), and Canada (6.1%, n = 15). Eleven of the included articles were based on international studies involving several countries, ranging from six to 44 involved, with nine of these eleven articles stemming from the same (HBSC) study [22]. The results reveal that while the reviewed articles originated from a total of 59 different countries, the United States had produced a vast majority of these and several of the countries had only produced one or two. Table 1 and Fig. 3 show the ten most frequently occurring countries of origin and the respective percentage out of the whole sample, visualized in two different ways.

Country of origin

Full size image

A total of 59 countries of origin were coded for. This table show the ten most frequent countries and their respective percentages out of the total 244 articles.

In terms of article type, 95.2% (n = 237) of the reviewed articles had been published in scientific journals and the remaining 4.8% (n = 12) were published in peer-reviewed conference proceedings. Out of the 244 reviewed articles, a vast majority (91.4%, n = 223) were based on empirical data while the remaining 8.6% (n = 21) were review articles (see Table 2).

71.3% (n = 174) out of the total 244 reviewed articles were based on studies conducted taking a quantitative approach, 30 (12.3%) were based on qualitative studies and 7.8% (n = 19) had used mixed methods. The remaining 8.6% (n = 21) of the included records consisted of various kinds of review articles. In addition, 57.4% (n = 140) out of the 244 articles were coded for being cross-sectional studies. In 103 of those articles, the authors themselves explicitly stated a cross-sectional design, while in 37 of the articles, the cross-sectional design was not explicitly stated, yet implicitly there and thereby coded for. Only 16% (n = 39) of the reviewed articles were based on longitudinal studies (Table 3).

In relation to the broader methodological approach, articles were additionally coded for type of method used for data collection. Given that methodological variables are not mutually exclusive, articles were coded for multiple methods when applicable, why the total number of methods used in Table 4 exceeds the total number of included studies.

What is especially clear from Table 4, is that out of the 223 empirical articles (21 review articles excluded), a vast majority (86.9%, n = 212) relied upon data collected through users’ self-reports of their use of social media. Most of the articles were entirely based on different kinds of self-report data, and only three of the reviewed empirical articles (1.3%) relied solely upon user study data with all three of those involving observations at participatory design workshops and similar researcher-initiated activities [23, 24] and one involving additional analysis of mobile application log data as well [25]. In addition to articles relying solely on self-report data and articles not involving self-report data at all, a few of the empirical articles (35.9%, n = 8) involved both. An approach identified within three of these articles was that of collecting data through a combination of participant observations of young users during researcher-initiated activities with semi-structured individual interviews [26, 27] or focus group interviews [28]. Additional approaches were found in [29] who combined data from field experiments with survey data and in [30] who conducted an experiment where log data from use of a social media applications was combined with semi-structured interview data [30]. With a stated purpose of obtaining “not only narratives but also insights into practices and forms of media use” [31, p. 3], Kutscher and Kreß complemented their interview data by inviting the young participants of their study to show their Facebook profile or smartphone during interviews to add stimuli to the conversations. In a study by Bowman-Smith et al. (2021), participants completed self-report questionnaires with individual sessions where they viewed pictures of other adolescents on a simulated social media app and choose between pre-written comments to send to a recipient [32]. Combining not only different methods and data, but also bringing different perspectives together, Kiefner-Burmeister et al. (2023) assessed generation Z’s social media use through parent reports, young users’ self-reports and log data of social media application use data from their smartphones [33].

Age varied among participants involved in the reviewed studies, ranging from 0 to 25, with most studies involving adolescents aged 10–17. While this wide range and heterogeneity makes comparisons between the studies difficult, it is however obvious from the papers reviewed that there is still a lack of studies examining social media use among the youngest children, which is in line with previous work making similar observations [26]. Apart from age, the stakeholder perspectives that the participants represented was also analyzed and is visualized in Table 5. Among the included 244 articles, a majority of 73.4% (n = 179) involved children and adolescents themselves and offered insights derived from data collected to understand their perspectives on social media use. 5.3% (n = 13) of the 244 records involved individual parents and the parental perspective. In addition to these studies where children, adolescents and parents were involved individually, 7% of the studies (n = 17) involved data collected from parent–child dyads where a child, or adolescent, was involved together with their parent, thereby providing dual perspectives on social media use. In one study for example, Daneels and Vanwynsberghe (2017) conducted a qualitative interview study where parents and adolescents were initially interviewed together, and thereafter separated to be interviewed individually as well [34]. Other similar perspective-combining approaches were taken in studies who involved additional family members (2%, n = 5). Four additional papers (1.6%) combined perspectives of both parents and minors as well, yet did not rely on data from related individuals stemming from the same family. A few additional studies (2%, n = 5) involved professionals working with children and adolescents and provided insights from the perspectives of healthcare personnel (such as doctors and youth mental health service providers) and educational staff (such as teachers and school support service staff). 8.6% (n = 21) of the articles were review articles, therefore building upon a variety of original data involving minors.

In terms of distribution among disciplines, most of the reviewed articles had been produced by scholars within either the medicine and health or the psychology disciplines (see Table 6). As many as 38.9% (n = 959) out of the total 244 records were published in medicine and health and the second most productive discipline, psychology, had produced 24.2% of the reviewed papers (n = 59). Far behind, and even despite specifically searching for eligible studies within the ACM Digital Library, computer science articles constituted only 5.3% (n = 13) of the included papers. While both of the most well-represented disciplines, medicine and health and psychology, as disciplines include numerous subfields of studies, articles stemming from these disciplines were also coded with subcodes to indicate subfield of study. Out of the 97 articles coded as belonging to the medicine and health discipline, the five most frequent subcodes were: psychiatry (n = 22), public health (n = 21), pediatrics (n = 19), nursing (n = 6), and sport medicine (n = 3). As a great number of the included articles were also conducted within psychology, these articles were also sub coded, reveling a most frequent subcategory of paedology (n = 13).

To better understand the discourse on social media use among children and adolescents, author keywords, provided in 223 out of the 244 reviewed papers, yielding a total of 2355 individual words, were analyzed. While some of the words were used in various combinations, analyzing the most frequently used author keywords individually provided an interesting insight into the discourse surrounding the studies made on social media use among children and adolescents. While 353 words only occurred once, the twenty most frequently used words within the keywords are listed in appendix Table 7. Less surprisingly, considering that the majority of the included papers stemmed from the medicine and psychology disciplines, several of the most frequently used keywords are related to health. When generating a word cloud out of the total 2355 individually used words in the keyword’s sections, the pattern was visualized further (see Fig. 4).

Visualization of author keywords

Full size image

Results from scoping reviews are usually analyzed both numerically as well as thematically [12, 13]. For the previous section, the numerical analysis was presented through a mapping and visualization of the quantifiable results, together painting a picture of a rather skewed body of previous work. For this following section, the findings are discussed further by unpacking the three most prominent themes that emerged from the thematical analysis. The three themes include geographical and disciplinary skewness, methodological constraints and lack in participant diversity.

4.1 Theme 1: Geographical and disciplinary skewness

The results show that there has been a steady increase in peer-reviewed publications on social media use among children and adolescents produced over the last five years, indicating a steadily growing body of work. From the results presented however, it is clear that the results stem from a limited set of producers, both geographically and discipline-wise. While the 59 countries of origin coded for suggest diversity, a vast majority of the papers included in this scoping review stemmed from Western, English-speaking countries, with by far the most of those produced in the United States. In an ideal world, research findings are judged simply by their quality. While this is desirable, previous work has established that the perception of a paper’s quality and relevance tend to be affected by its country of origin [35]. Also, another potential danger of this concentration is that the research agenda may become skewed towards these countries’ interests, leading to an imbalance in global knowledge production and ignoring of important issues relevant to less represented regions.

Another central concern with the reviewed research on social media use among children and adolescents is that it has primarily been produced by scholars within the health-oriented disciplines. The health-orientation of the reviewed papers is also indicated by the analysis of author keywords visualized in Table 7 and Fig. 4. While the results from articles produced by scholars within health-oriented fields have brought attention to the possible impacts of social media use on childrens and adolescents’ mental and physical health, with studies exploring the relationship between social media use and sleeping patterns (such as in [36, 36,37,38,39,40,41,42,43] and [44]), attention deficit hyperactivity disorder (ADHD) symptoms (such as in [45, 46]) and anxiety and depression [21, 47,48,49,50], this perspective and the common objectives, research questions and findings of these studies only covers some aspects of what constitutes social media use and the impact it might have on user health and well-being. To fully understand the impact social media use might have on children’s and adolescents’ health and well-being, we must first strive to gain a fuller understanding of their situated social media use, the social practices involved and the meaning these practices have to them, which most certainly require additional studies from a additional disciplines. Also, future work should explore whether previous findings apply to non-Western contexts.

4.2 Theme 2: Methodological constraints

A concern related to the first theme is the uniformity in methodological approaches and constraints in terms of distribution of methods. Different disciplines rely on different traditions in terms of methodological approaches, and the limitations in distribution among different disciplines might therefore be a partial explanation to the homogeneity in methods used within the reviewed papers. As highlighted in the numerical analysis (Tables 3, 4), a majority of the reviewed papers were based on studies conducted taking a quantitative approach and relied upon data collected through self-report surveys and questionnaires. Although self-reports are acknowledged to be effective in terms of gaining insights otherwise difficult to obtain directly from users themselves instead of from third parties, such as parents [51], it is pointed out in several of the papers reviewed that relying upon self-reports of social media use also comes with limitations [41, 52,53,54,55]) as they, as pointed out by Roberts et al. (2023) only “represent what participants are willing to share” rather than their actual and full experiences [56, p. 218]. All methods come with limitations and weaknesses, yet previous work has showed that self-reported media usage does not constitute a great measurement when aiming to understand social media use due to its fragmented and ubiquitous nature, making the accuracy of self-reported measures weak [57]. Also, as pointed out by Anthony (2023) “adolescents, in particular, are poor reporters” [58, p. 9] as they tend to overestimate their time spent on social media [59].

Another methodological constraint visible and in the reviewed papers is the reliance on cross-sectional studies. As acknowledged by several authors, this methodological conduct comes with limitations such as lacking possibilities of drawing insights on changes and variations over time and to establish causal relationships (see for instance [60,61,62,63,64,65,66,67] and [68]). As put by Bloemen and De Coninck [60], cross-sectional data “merely reflect snapshots of adolescents in time” [60, p. 9]. Among the reviewed papers, the majority build upon cross-sectional studies and only a very limited number of the reviewed papers (16%, n = 39) were based on longitudinal studies. However, the need for further longitudinal research is pointed out in several of the reviewed papers (see for instance [58, 69,70,71,72,73]).

The results reveal that only 12.3% (n = 30) of the reviewed papers were based on qualitative work solely, with only three papers involving participant observations. Ethnographic work, such as participant observations [74], can offer a profound understanding of social media practices in their natural contexts. This approach allows researchers to witness interactions and social practices in real time, providing insights that go beyond self-reported data. By employing ethnographic methods, future studies can address the current gaps in understanding the complex and situated nature of social media use among younger populations. While the qualitative studies in this scoping review were few, additional qualitative work on social media use among children and adolescents is suggested within several of the reviewed papers (see for instance [26, 60, 75,76,77,78]). For, Siongers and Spruyt (2024) suggest that “further qualitative research that tracks young individuals over time and closely studies the interplay between their offline and online social interactions would enable us to gain a deeper understanding of the emergence of problematic social media use and the crucial factors involved in this process” [77, p. 191]. Twigg et al. (2020) also acknowledge that there is still a need for a deeper understanding of how social media technology is being used [78]. A way to move beyond possible self-report biases and receiving only “what people are willing to share” [56, p. 218] and towards a deeper understanding of how social media is being used, is to design future studies of children’s and adolescent’s use of social media to approach the matter through methodological manners where the aim is to observe and analyze use ‘in-situ’, for example through participant observations of situated social media use, possibly as suggested by Lawrence et al. (2022), collecting data through wearable cameras as capturing tools for an even more objective measure [51]. Also, additional mixed methods work involving objective social media data, is argued for [79].

From this scoping review, it is clear that little attention has been paid to studying and understanding the situated use of social media among children and adolescents as it plays out in its natural occurrence. While data collected through surveys, questionnaires, and similar quantitative approaches is undeniably superior in terms of revealing patterns within big data sets and large populations and allowing for replicability and generalizability to a significantly greater extent than qualitative approaches, quantitative studies do not adequately capture the details of the complex and varied nature of social media use.

4.3 Theme 3: Lack in participant diversity

Adolescents are overrepresented in previous research on social media use among children and adolescents, and there is a lack of studies involving the very youngest users (three to eight years old) [26]. Possible explanations for this meager body of work may be the age restriction of thirteen for registering accounts on social media platforms which, despite studies concluding that younger users find ways to create accounts for themselves even when they have not reached the required age, makes it difficult to map the use among these young users. Another possible explanation for the lack of work involving minors are ethical challenges that comes with conducting research of ICT use among minors and lack of ethical guidelines for doing so [80]. While there is an established tradition of following the continuously evolving and adapting ethical guidelines developed by the Association of Internet Research (AoIR) [81,82,83] when conducting research of social media use among adults, it has been established that in terms of minors, the literature is still underdeveloped [80]. While involving the youngest users and prioritizing children’s perspectives is intuitively appealing, researchers also have a significant responsibility to protect these young participants from any potential harm caused by the research process. This responsibility is especially crucial when including minors, requiring careful measures to ensure their safety during data collection. While participatory research or ‘being there live’ [84] could be argued for when aiming to get a deeper understanding of how social media technologies are being used in practice, when conducting studies of children in online, non-physical spaces, such as during their interaction in social media, there are additional ethical considerations to address. For instance, one central ethical dilemma when collecting publicly available data from the online, non-physical spaces that social media provide is whether this data is fair to use without informing the authors and creators behind it. Although data may have been made publicly available through social media, for instance through posting videos on TikTok and selfies on Instagram, it is often the case that the creators may not have considered all possible uses of the data and may feel a strong sense of ownership of it [85]. While it is not possible to fully foresee how the participants might reflect on their contributions in ten, twenty or even fifty years from now, considerations on the long-term impact of presenting their content created in a different context should be made carefully. For example, observations can be made overt, i.e. the participants will know that they are being researched and participation will be confirmed through signing informed consent forms [85].

Among the 244 articles reviewed, a vast majority of 73.4% (n = 179) involved children and adolescents themselves and offered insights derived from data collected to understand their own perspectives on social media use. However, an interesting point made by Sun et al. (2020) is that a large part of previous work investigating the effects of social media on young children have studied children’s social media activities “from an adult’s point of view, without much consideration for what children find interesting and useful” [28, p. 332]. This might suggest that even though children and adolescents are involved in studies, questionnaires, survey questions and interview guides are prepared from the adult researcher’s point of view, possibly lacking in terms of understanding the context in which social media is being used. A possible way to address ethical concerns and protect children and adolescents while also bringing in these young users’ perspectives could be through data collection with parent–child dyads. 7% of the reviewed studies (n = 17) involved data collected from parent–child dyads where a child, or adolescent, was involved together with their parent, thereby providing dual perspectives on social media use (see for instance [67, 86,87,88] and [89]). Family-centered qualitative investigations are argued for as one way of addressing the lack of studies involving children [26].

Participant diversity is crucial in future research because, even when participants share demographics such as age, sex, and country of origin, social media use and activities remain highly heterogeneous at the individual level [90, 91]. These differences in use, expectations and interactions must be considered when recruiting informants, designing interview guides and collecting and analyzing data. Understanding the complexity and nuances of social media use within varied user groups at a micro-level is essential for effectively informing policy, practice, and future design.

This scoping review offers a comprehensive overview of trends and gaps in existing peer-reviewed research on children’s and adolescents’ social media use. The review reveals a growing interest in this topic, implied by the steady increase in studies over the past five years, but also highlights a need for greater geographical and disciplinary diversity, varied methodological approaches, and more heterogeneous participant involvement in future research.

The generation of children who grow up today live in a “permanently online, permanently connected” world [92]. Through their time spent with contemporary ICT’s, they grow up to become technology-savvy adult citizens, prepared to live, work, and socialize in a digitized future, yet there are also concerns raised regarding what impact screen time and social media use might have on their coming of age. Developing research-based knowledge on this generation’s screen time practices is crucial in addressing the potential consequences of technology-mediated interactions. Therefore, it is necessary to incorporate diverse data, methods, and research approaches to better understand the complex dynamics and practices involved in social media use, especially among children and adolescents. The significant increase in research on social media and smartphone use among children and adolescents in recent years can be partly explained by applying the lens of moral panic theory [93]. This sociological concept describes how heightened societal fears about new technologies can influence research agendas, leading to an emphasis on potential negative impacts [93]. Similar to historical instances with video games and television, the rapid adoption of social media and smartphones has prompted concerns about their effects on young users, possibly driving a wave of studies aimed at addressing these anxieties. While it is crucial to investigate these potential impacts, it is equally important to recognize the pressure that moral panic can place on researchers to confirm pre-existing societal fears [94]. Therefore, we must strive for a balanced research approach that considers both the benefits and challenges of technology use, utilizing diverse methodologies to provide a comprehensive understanding of its effects. By doing so, we can ensure that research outcomes contribute meaningfully to the public discourse and inform research-based design, policy and practice.

The findings of this scoping review highlight a significant imbalance between the number of qualitative and quantitative studies, with quantitative, cross-sectional studies relying on self-reported questionnaire results being the most common. However, quantitative studies do not adequately capture the details of the complex and varied nature of social media use, nor do the common cross-sectional nature of those studies enable insights to be drawn on changes and variations over time or establish causal relationships. This review suggests a pressing need for more empirical, qualitative research that examines various aspects of situated social media use from the perspectives of children and adolescents, especially from the youngest users. Key aspects such as the activities and practices that constitute social media use from minors’ perspectives, how children and adolescents objectively use social media, the significance these technologies hold for them, and how interacting with and through these platforms affects their lives still require further exploration. A deeper understanding of these factors is essential to fully grasp the interplay between social media use and children’s and adolescents’ development, learning, and social life. Moreover, additional and carefully designed longitudinal studies are necessary to gain causal explanations and a deeper understanding of the evolving dynamics of social media use among younger populations. While longitudinal studies are useful for exploring potential trends and associations, they remain limited in establishing causality due to their correlational nature. To fully assess causal relationships, it is crucial to integrate longitudinal data with additional methods. Notably, recent longitudinal studies, such as those by Heffer et al. [95], Jensen et al. [96] and Steinsbekk et al. [97] underscore the complexity of isolating causal effects within social media research.

Social media use does not occur in a social vacuum; thus, understanding the broader social context in which it takes place is crucial. To accurately assess the impact of social media use on children and adolescents, it is essential that future studies are designed to account for additional influential factors, such as family and friendship dynamics, school-related and academic stress, and prior mental health conditions. Such study designs are crucial to advance our understanding of the complex interplay between social media use and its impact on young people’s lives. From the results from this scoping review, additional qualitative work, such as ethnographic fieldwork and participant observations in physical and non-physical settings, conducted to gain a comprehensive understanding of situated social media use, is encouraged. However, it is essential to address ethical concerns and develop guidelines for researching children’s and adolescents’ mobile technology-mediated interactions in both physical and non-physical spaces. This will enable secure involvement of young users in future research while also emphasizing their agency.

No studies are without limitations, including this one. Scoping reviews, by nature, provide a narrative account of available research rather than evaluating the quality of evidence [13]. This scoping review also has specific limitations, such as potential gaps in capturing the full extent of social media usage and user experiences among children and adolescents. Also, while the geographical skewness identified through this scoping review is strong, it is important to acknowledge that the geographical skew towards Western countries might be a possible consequence of the inclusion criteria requiring articles to be written in English. While this criterion ensured accessibility, it inadvertently limited the review to English-speaking research, potentially excluding significant studies published in other languages. However, this limitation highlights the need for future reviews to incorporate translated papers and engage multilingual research teams to better capture global research efforts on social media use. Moreover, we must acknowledge the resource disparity that often favors Western institutions, suggesting a need for collaborative international efforts to bridge this gap. By expanding our linguistic and geographic scope, we can achieve a more balanced understanding of social media’s impact across diverse populations, cultures and regions. In reference to the included review papers in general, and meta-analysis papers in specific, it should be noted that the quality of the findings of those is determined by the quality of studies they incorporate [4]. For instance, Ferguson et al. [4] concluded from a meta-analysis of social media experiments that these experiments had extensive limitations, such as lack of preregistration practice and only few reliability checks, and therefor limitations in evidence for causal effects they are able to provide. Despite the study limitations, this scoping review reveals significant knowledge gaps within the research field and multiple opportunities for future research.