Messaging, Posting, and Browsing: A Mobile Experience Sampling Study Investigating Youth’s Social Media Use, Affective Well-Being, and Loneliness

Employing a mobile experience sampling design, we investigated in the present study how different types of mobile social media use relate to young individuals’ momentary affective well-being and momentary loneliness. We differentiated between three types of social media use: Messaging, posting, and browsing. Moreover, we studied fear of missing out (FoMO) as a moderating variable. We collected data from 79 middle and late adolescents (M age = 17.55 years, SD = 1.29; 59% girls) yielding 956 momentary assessments. The results showed that messaging and posting were positively related to affective well-being, while browsing was associated with higher levels of loneliness. Furthermore, some of the relations between social media use, affective well-being, and loneliness were also moderated by FoMO. Our results highlight the need to differentiate between different types of social media use, to include individual predispositions, and to apply methods that account for daily fluctuations in psychological well-being when studying the complex relationship between youth’s mobile social media use and well-being.


Introduction
Due to the ubiquitous availability of mobile devices and the increased accessibility to social media platforms, teens are "permanently online" (Vorderer et al., 2016). In particular, mobile instant messaging (IM) services and mobile social networking sites (SNS)-hereinafter referred to as social media-play a crucial role in youth's daily lives to maintain and create social interactions with their peers (Rideout & Robb, 2018). For example, 92% of 11-to 18-year-old Austrians possess a smartphone (Education Group, 2019), with WhatsApp, YouTube, and Instagram representing the most popular social media platforms (SaferInternet, 2020).
By integrating media production, distribution, and consumption on one device, mobile social media provide an added layer of connectivity in social life (Campbell, 2019). In particular, social media lend themselves well to account for young individuals' age-specific recreational, educational, and social needs (Konijn et al., 2015;Mascheroni & Vincent, 2016). Friends and peers can be easily accessed as messaging services allow for instant communication via texts, pictures, voice messages, or videos while on the go. The smartphone and its affordances to digitally document and express thoughts and experiences allow its users to present and experiment with their identity. Likewise, the posts of peers, celebrities, and other people can be browsed, commented on, and shared, for example, while standing in line or while waiting until the pedestrian lights turn green.
It is important to study youth's social media use because the period from early to late adolescence is characterized by fast developmental and behavioral changes, which determine psychological well-being in later life (Wilson & Wilson, 2014). Given the immense popularity of smartphones and communication apps, research investigating the potential beneficial and harmful effects of these technologies on youth's well-being increased in recent years. Accumulated research on the relationship between social media use and youth's well-being is complex, leading to conflicting results of positive, negative, and nonsignificant outcomes on well-being (Orben, 2020;Schemer et al., 2021). Although the findings from this research provided valuable insights, they come with important methodological limitations, and at least four research gaps can be identified.
First, given that the research field is shaped by cross-sectional survey designs, most studies have not disentangled within-from between-person associations of social media use and wellbeing (Beyens et al., 2020;Orben et al., 2019;Schemer et al., 2021). While more recently, researchers have implemented longitudinal survey designs to overcome this issue, most analyses focused on between-person associations. Between-person associations help understand whether individuals using social media more or less often than their peers experience lower or higher levels of well-being than these peers. However, these findings cannot be translated to answer the question of whether an increase in social media use is associated with an increase in well-being. To make such statements, within-person analyses are needed, and such analyses require a minimum of three measurement points across time.
Second, measures and research designs that reflect momentary states of well-being and loneliness are still missing (Meier & Reinecke, 2020). Most findings rely on trait level assessments, while many psychological processes, such as affective well-being, fluctuate with daily experiences across youth (Larson et al., 2002). Specifically, the transitional age period between early adolescent years and late adolescence is marked by instabilities in positive emotions (Larson et al., 2002). The present study focuses on affective well-being and loneliness, representing an essential component of subjective well-being (Diener et al., 2018). Affective well-being, in particular, has been insufficiently studied in this area of research (cf. Beyens et al., 2020;Weinstein, 2017).
Third, despite important efforts to provide further conceptual depth to research on social media use exists, in most studies, general social media use or single aspects of social media use (e.g., Facebook use) have been examined. Empirical findings showed that different activities, such as browsing content (also referred to as broadcast communication), are more common than social or interactive activities when using the smartphone (Elhai et al., 2016) or SNS (J. L. . Such a distinction of types of use is essential because different communication use patterns have been associated in different ways with well-being (Goodman-Deane et al., 2016;Liu et al., 2019). For example, while posting content about oneself on social media has been associated with higher levels of happiness and self-esteem, content consumption has been negatively related to well-being . Moreover, researchers have classified smartphone use along two dimensions: active-passive and private-public (Frison & Eggermont, 2020;Verduyn et al., 2015;K. Wang et al., 2018). Within these dimensions, messaging has been labeled as private active type of use, posting content as public active, and scrolling through the social media feed as public passive use (e.g., Frison & Eggermont, 2020;Verduyn et al., 2015). Other researchers relied on the distinction of communicative and non-communicative uses (Chan, 2015;Stevic et al., 2021) (Stevic et al., 2021) to explain relationships with individual's well-being.
Finally, fear of missing out has not been sufficiently investigated in previous studies despite the suggestion that it is essential for a thorough understanding of the relationship between young individual's smartphone use, loneliness, and well-being (Przybylski et al., 2013). Youth is marked by profound psychological and physiological developments involving a gradual detachment from the parents and an orientation towards peers (Wilson & Wilson, 2014). Against this background, we introduce fear of missing out (FoMO) (Przybylski et al., 2013) as a moderator explaining the relationship between young individuals' types of mobile social media use (i.e., messaging, posting, and browsing), affective well-being, and loneliness.
The present study aims at addressing these research gaps by investigating the associations between different types of daily SNS use and youth's momentary affective well-being and momentary loneliness with a mobile experience sampling design (MES). The experience sampling method allows participants to repeatedly report their momentary behavior, feelings, and cognition, thus providing more precise measures than traditional survey designs (Naab et al., 2018). More specifically, we investigated the associations of active private (i.e., messaging), active public (i.e., posting), and passive public (i.e., browsing), and youth's affective well-being and momentary loneliness. We included fear of missing out as a moderating variable in this relationship.

Types of Social Media Use and Youth's Affective Well-Being
Drawing on media richness theory (Daft & Lengel, 1986;Sheer & Chen, 2004) and social presence theory (Short et al., 1976) and prior empirical work, we examine different types of social media use and digital communication (Burke & Kraut, 2016;Chan, 2018;Chan & Li, 2020;Goodman-Deane et al., 2016;Liu et al., 2019;Schmuck et al., 2019;Stevic et al., 2021). Media richness theory posits that communication channels vary in their capacity to transmit detailed and timely information (Daft & Lengel, 1986;Sheer & Chen, 2004). According to this theory, rich media (e.g., phone calls or video calls) are characterized by communicative intimacy and immediacy and thus allow for more efficient communication and problem-solving than leaner media (e.g., email). Moreover, richer media also allow for more sensory cues (e.g., facial expression and tone of voice), compared to leaner media. Similarly, social presence theory (Short et al., 1976) relies on the assumption that media channels differ in their capability for creating social presence and social proximity. Within this framework, media differ on how much social cues they can convey. Rich media, as compared to lean media, can carry more personal information and allow for more social presence-arguably, such communication characteristics are valued in close relationships. These theoretical assumptions have been applied to explain why different types of digital media are associated with better or poorer overall well-being (Chan & Li, 2020;Liu et al., 2019), social connectedness (e.g., Nguyen et al., 2021), or relationship quality (Chan, 2018;Chan & Li, 2020;Stevic et al., 2021). Although media richness theory and social presence theory distinguish between different media channels and not media uses, its core tenets can be applied to underpin the selection of different types of social media and to substantiate-in conjunction with further theoretical frameworks-our hypotheses and research questions regarding affective wellbeing and loneliness (see also, Chan & Li, 2020).
With regards to social networking site use specifically, Liu et al. (2019) differentiate between four types of use in explaining better or poorer overall well-being: interaction (i.e., replying, commenting, and liking), self-presentation (i.e., status updating and photo posting), entertainment (gaming and entertainment) and content consumption (i.e., browsing, searching, and monitoring). In our study, we used this differentiation combined with previous conceptualizations (i.e., activepassive, private-public) to study three aspects of social media use: Sending messages as an active private type of use, posting as a type of active public self-presentation, and browsing as a type of passive public content consumption.
Drawing on media richness theory (Daft & Lengel, 1986), sending messages can be considered as rather lean media type when compared to voice calls. Common messengers (e.g., WhatsApp) allow for sending messages in various modalities (including text, voice, or video) which is typically followed by asynchronous feedback, while voice calls enable immediate feedback. However, when compared to posting content for a wider audience or browsing through a social media feed, sending private messages (which is typically done among close relationships) likely elicits more social presence compared to the other two media types. Also, browsing does not entail feedback, thus, messaging, posting, and browsing can also be classified in descending order in terms of their richness and their capacity to provide social presence.
Although systematic research on different types of smartphone use with youth samples is mostly missing, preliminary findings on different SNS use patterns provided valuable insights into its effects on the user's psychological well-being. In what follows, we present the three types of use in relation to affective well-being.

Instant Messaging
Mobile instant messaging (IM) apps allow for private conversations with individuals or groups, such as Snapchat, WhatsApp, or WeChat. Such apps allow the exchange of text messages and other types of communication, including photos, videos, and audio files. In Austria, where we conducted the study, the messaging services WhatsApp (85%) and Snapchat (59%) were among the most popular social media platforms among youth aged 11-17 years (SaferInternet, 2020).
Typically, instant messaging services are used to exchange text messages and photos with close friends and family members (e.g., Piwek & Joinson, 2016;Valkenburg & Peter, 2007). Therefore, messaging has been considered as an active private type of social media use , 2020. Mobile instant messaging services, such as WhatsApp, can be conveniently adapted for varying levels of richness and social presence (Cui, 2015)-ranging from lean text messages to rich audiovisual calls. In the Internet-enhanced intimate selfdisclosure hypothesis, Valkenburg and Peter (2009) propose that IM results in enhanced intimate self-disclosure, which is beneficial for youth's psychological well-being. In line with this hypothesis, we assumed that young individuals benefit from IM because it is likely to promote connection and belongingness-which are crucial predictors for social development and affective well-being (Clark et al., 2018;Wilson & Wilson, 2014). In a previous study, general messaging use (e.g., WhatsApp, WeChat) to communicate with friends has been positively associated with various indicators of social well-being, including to social support and friendship satisfaction among adults (Chan, 2018) increased friendship closeness among adolescents (Pouwels et al., 2021;Valkenburg & Peter, 2009), and positive changes in emotional states among youth (Dolev-Cohen & Barak, 2013). The study findings by Chan and Li (2020) among Chinese smartphone users further substantiate this claim. Mobile WeChat provides an asynchronous tool to communicate with friends if social support is needed, which enriches existing relationships and, eventually, result in psychological well-being (Chan & Li, 2020).
However, these studies did not differentiate between passive and active Instant Messaging use or within-person associations. There is only one study we are aware of, investigating withinperson associations of messaging (i.e., active and passive WhatsApp use) and affective well-being (Beyens et al., 2020). Interestingly, and in contrast to previous findings, the findings showed a positive association between passive WhatsApp use and adolescents' affective well-being. No significant associations were found between active WhatsApp use and affective well-being. Thus, research on the relationship between messaging and affective well-being is somewhat mixed. However, media richness theory (Daft & Lengel, 1986;Sheer & Chen, 2004) and the Internetenhanced intimate self-disclosure hypothesis (Valkenburg & Peter, 2009) raise the expectation of a positive correlation between the two constructs. Therefore, we proposed the following hypothesis: H1: Messaging will be positively associated with youths' momentary affective well-being.

Posting Content
Posting content on SNS is defined as a self-presentation type of SNS use , also referred to as active public social media use (Frison & Eggermont, 2020;K. Wang et al., 2018).
With the smartphone readily available, social media users can post almost any type of content at any given time or place. This behavior lends itself well to fulfill the developmental need of identity-seeking experiences (Konijn et al., 2015). Self-presentation behavior online is tied to certain expectancies, such as receiving attention or stimulating communication with others (e.g., Sung et al., 2016). Usually, these expectations are met given that young social media users predominantly receive positive feedback on their profiles (Meeus et al., 2019;Valkenburg et al., 2006). Empirical evidence from the studies by Eggermont (2016, 2020) supports our reasoning. Based on panel survey data, the authors found that active public Facebook use increased positive public feedback, which in turn, was positively associated with adolescents' perception of friend support (Frison & Eggermont, 2020). Similarly, adolescents' active public Facebook use was associated with perceived online support . A recent study linking Facebook log-data with self-reported affective well-being further substantiates our rationale, as receiving likes and comments to self-presentation on social media has been positively associated with young adults' affective well-being (Marengo et al., 2021). Capitalization offers another explanation for this association. Capitalization refers to the ability to enhance positive affect by holding on to these moments by sharing them with others (Peters et al., 2018). Given that users typically post selected, positive content online, such capitalization processes are likely to occur as positive self-presentation is likely to induce (positive) reactions from others, which has been positively associated with affective well-being among Facebook users (Kim & Lee, 2011). These outlined mechanisms apply primarily when posting positive or authentic content. However, it should be acknowledged some young social media users experience constant pressure to present an idealized version of themselves (Rideout & Robb, 2018), and others even provide inauthentic or false online self-presentation (Twomey & O'Reilly, 2017). Informed by literature on strategic self-presentation (Goffman, 1956), we assume that publicly posting content on social media-independent of its authenticity-is associated with certain motivations (e.g., receiving attention and social approval) and that these motivations are usually met by immediate gratifications, for instance, via likes and comments (Bareket-Bojmel et al., 2016;Kim & Lee, 2011;Marengo et al., 2021). Even though there might be negative effects of posting inauthentic content in the long run, we assume that specifically on a momentary level, posting content is likely to result in short-term gratifications, which are positively associated with higher levels of momentary affective well-being.
H2: Posting will be positively associated with youths' momentary affective well-being.

Browsing
Descriptive research suggests that passive public or non-communicative activities, such as browsing content on social media, are more common than social or communicative activities when using the smartphone or SNS (Beyens et al., 2020;Verduyn et al., 2015). In comparison to messaging and posting, browsing can be considered as a lean type of media use that creates considerably less social presence. Similarly, while messaging and posting are considered active types of social media use, browsing or scrolling through news feeds, looking at friends' pages and pictures have also been referred to as passive public or non-communicative social media use. Drawing on social comparison theory (Festinger, 1954), researchers have suggested that passive social media use elicits upward social comparison and envy, which subsequently result in decreases in well-being (e.g., Schmuck et al., 2019). However, the empirical evidence on that hypothesis is inconclusive. Among adult participants, some researchers found a negative association between passive use and well-being (Burke et al., 2010), whereas others found no association between passive social media use and well-being (Burke & Kraut, 2016;Verduyn et al., 2015), and some even found more positive than negative emotions while browsing Facebook (Lin & Utz, 2015).
Regarding the evidence on affective well-being among youth, the empirical evidence is even further limited. No main effect of browsing on affective adolescents' well-being was found (Weinstein, 2017). Results derived from an experience sampling study among adolescents further challenge the passive social media use hypothesis (Valkenburg et al., 2022). On the betweenperson level, passive public social media use (i.e., browsing) was negatively associated with affective well-being. On the within-person level, however, there was no significant overall effect. Given the inconsistent findings, scholars recently challenged the hypothesis of passive use being detrimental to youth's well-being (Valkenburg et al., 2022). Thus, we refrained from proposing a hypothesis and formulated the following research question instead: RQ1: How will browsing be associated with youth's momentary affective well-being?

Types of Social Media Use and Youth's Loneliness
Loneliness can be defined as the subjective lack of social interactions and is a central predictor of subjective well-being (Cacioppo et al., 2006). As an indicator of ill-being, it is important to note that loneliness should not be understood as simply the opposite of well-being (Meier & Reinecke, 2018;Valkenburg et al., 2021). Rather, meta-analytical evidence suggests that social media use is associated with adolescents' well-being and ill-being , which justifies looking at loneliness as a dependent variable in addition to affective well-being. Although a considerable amount of research was dedicated to answering the question of whether social media use causes or predict loneliness (Song et al., 2014), longitudinal research on different types of social media use and its relationship to momentary loneliness is still needed and, thus, deserves further scholarly attention. Based on media richness theory (Daft & Lengel, 1986;Sheer & Chen, 2004) and social presence theory (Short et al., 1976), we assumed that active types of social media use (i.e., messaging and posting) would provide substantially more social intimacy and proximity than browsing. Thus, we assumed that these types of social media use would likely reduce momentary feelings of loneliness. In what follows, the literature on the relationship between instant messaging, posting content, and browsing and loneliness are summarized. Given that the empirical evidence on these relationships is inconclusive, we formulate three research questions.

Instant Messaging, Posting Content, and Browsing
As detailed above, the Internet-enhanced intimate self-disclosure hypothesis (Valkenburg & Peter, 2009) suggests that young individuals use messenger services primarily to communicate with existing friends-resulting in enhanced intimate self-disclosure and higher levels of social wellbeing. Thus far, scholars provided only limited empirical evidence for this hypothesis. For instance, socializing on SNS was linked to decreased loneliness among Spanish adolescents in a cross-sectional survey design (Apaolaza et al., 2013). However, this relationship did not hold up in a panel study with a representative sample (16+ years) from Germany (Dienlin et al., 2017).
With regard to posting, researchers have argued that interactive use and self-presentation, such as posting personal updates, can help individuals tolerate the lack of offline social interaction and therefore reduce loneliness (e.g., Luo & Hancock, 2020;Song et al., 2014), as messaging or posting content (and receiving reactions to them) makes one's social network more salient, stimulates perpetual connectivity with existing networks, increases social support, and reduces the feeling of loneliness (große Deters & Mehl, 2013;Luo & Hancock, 2020). For example, a crosssectional survey showed that active public use of Facebook among girls was positively linked to perceived online social support . Große Deters and Mehl (2013) conducted an experimental study with undergraduate students to investigate the effect of posting status updates on Facebook on loneliness. Participants who received the instruction to post more status updates felt less lonely than participants who did not receive such instruction. This effect on loneliness could be explained as posting updates helped to maintain connectedness to friends by sharing daily experiences and by letting friends take part in one's life. However, in a crosssectional survey design, posting content on Instagram was associated with higher loneliness among undergraduate students (Yang, 2016).
By definition, browsing content does not involve interpersonal exchange. This lean type of media use likely evokes only little social presence. With regard to browsing, current findings are mixed. For example, Verduyn et al. (2015) found no effect of browsing Facebook on loneliness in an experimental study with adults. In a correlational study among undergraduate students, however, Yang (2016) showed that browsing on Instagram was related to lower levels of loneliness among an undergraduate student sample.
In conclusion, the findings on various types of social media use on youth's loneliness are mixed-longitudinal approaches allowing to test for within-person associations and momentary measures of loneliness are virtually non-existent. Therefore, we propose the following research questions: RQ2: How will a) messaging, b) posting content, and c) browsing be associated with youth's momentary loneliness?

The Moderating Role of Youth's Fear of Missing Out
We introduce fear of missing out (FoMO) as a moderating variable to explain the boundary conditions of our postulated relationships between adolescents' social media use, affective wellbeing, and momentary loneliness. Przybylski et al. (2013) define FoMO as the profound concern that others might be having (more) rewarding experiences from which one is excluded, which results in the desire to stay continually connected with what others are doing. In the past, FoMO has been primarily investigated in the context of social media use.
Young individuals, especially in adolescence, face several developmental tasks, and one of these tasks is to develop a sense of their identity. As part of their identity formation process, individuals particularly rely on their peers and have a strong need to belong (Wilson & Wilson, 2014). Mobile social media can provide an excellent tool to satisfy this need, as one can quickly check what others are up to via social media (Konijn et al., 2015). In line with self-determination theory (SDT; Ryan & Deci, 2000), we argue that youth's need for popularity and need to belong, representing specific facets of relatedness, predict FoMO, and that youth with high FoMO have a strong need for connection with their friends via mobile social media (Beyens et al., 2016;Lee, 2009). It can be further speculated that for those young individuals with high FoMO, who satisfy their needs of relatedness by messaging, posting, and browsing on social media might increase affective-well-being and reduce loneliness. Preliminary evidence among young adults suggests that FoMO moderated the relationship between social media use and negative affect (Riordan et al., 2021). Moreover, among adolescents (aged 11-19) with high levels of FoMO, general social media use was positively related to subjective well-being (Chai et al., 2019). However, given the lack of consistent empirical findings supporting this argument, we formulated the following research questions: RQ3: Does FoMO moderate the relationship between mobile social media use (i.e., messaging, posting, browsing) and a) momentary affective well-being and b) momentary loneliness among youth? Figure 1 depicts our hypothesized model, including the hypotheses and research questions.

Participants and Procedure
A total of 79 middle and late adolescents (M age = 17.53 years, SD = 1.29, range: 15-21 years, 58% girls, 41% boys) participated in the study yielding 956 momentary assessments (participation rate: 80.7%). We recruited the students from three schools in Austria in Spring 2018. All students who had received parental consent were allowed to participate in the study. Before starting the study, the researchers visited the students in school, informed them about the study goals and procedure, and asked them for active consent. Participants were also informed that they could participate in a lottery to win 25 vouchers worth 10 € each if they regularly participated in the mobile experience sampling study. First, participants received a link in class, which directed them to an online questionnaire assessing the moderator variable FoMO, demographics, and all statistical controls from the background survey (i.e., age, gender, school type, self-esteem, well-being related to friends, well-being related to family, and loneliness). To assess social media use (i.e., messaging, posting content, and browsing), momentary loneliness, and momentary affective well-being, we applied a mobile experience sampling procedure. We administered 15 short online surveys to each student for five consecutive weekdays. We prompted the participants at three different times throughout the day at 7 a.m., 3 p.m., and 8 p.m. After data collection, winners were drawn from participants who responded to at least 13 of 15 momentary assessments (70% of the participants). Our anonymized data and analysis scripts are available at https://osf.io/ htywp Measures MES Measures. Messaging, Posting, Browsing: Similar to a previous study on mobile online communication (Kross et al., 2013), we assessed the different types (i.e., active public, private public, and passive public) of mobile social media use with one item each. We measured mobile instant messaging by asking the participants to indicate how often in the past hour they used their smartphone to send messages, such as images, texts, or voice messages, e.g., via WhatsApp or Snapchat. To measure posting behavior, we asked participants how often, in the past hour, they have used their smartphone to post content (e.g., on Instagram). To measure browsing, we asked how often they have used their smartphone to scroll through posts of others (e.g., on Instagram). We asked the respondents to indicate their answer using a Likerttype scale including the options 1 = "never," 2 = "once," 3 = "several times," and 4 = "constantly" (messaging: M = 2.31, SD = 0.54; posting: M = 1.20, SD = 0.36, browsing: M = 1.97, SD = 0.63; between-person).
Momentary Affective Well-Being: In line with previous experience sampling studies (Beyens et al., 2020;Kross et al., 2013;Reissmann et al., 2018), we measured affective well-being with one item. At each assessment, we asked participants to indicate how they felt at the moment. We specified this question by asking participants to indicate their agreement on whether they felt happy. Participants answered on a 5-point Likert-type scale (1 = "strongly disagree"; 5 = "strongly agree"; M = 3.37, SD = 0.72, between-person).
Momentary Loneliness: Like previous experience sampling studies on social media use and loneliness (Kross et al., 2013;Reissmann et al., 2018), we used one item to assess momentary loneliness. As with momentary affective well-being, we asked participants to indicate whether they felt lonely at the moment using a 5-point Likert-type scale ranging from 1 = "strongly disagree" to 5 = "strongly agree" (M = 1.82, SD = 0.73, between-person). Measures from the Background Survey. Fear of Missing out (FoMO): To measure FoMO, we used two items based on the scale by Przybylski et al. (2013) ("I fear others have more rewarding experiences than me," "I get anxious when I don't know what my friends are up to," r = 0.19, p =.044, M = 1.99, SD = 0.82). We included only two items from the original 10-item scale for pragmatic reasons regarding the overall survey length. We selected those two items because they reflect the anxiety and fear aspects (as opposed to behavioral aspects) of the concept of Fear of Missing Out. Participants indicated their agreement with each statement with a 5-point Likert-type scale (5 = "strongly disagree"; 1 = "strongly agree"). We recoded the items for analysis so that higher values indicated higher FoMO.
Trait Loneliness: We assessed trait loneliness with the 8-item UCLA Loneliness Scale (Roberts et al., 1993), which has been shown to have good internal consistency among youth (Roberts et al., 1993). We asked participants to indicate their agreement with eight statements (exemplary item: "I feel isolated from others") on a 5-point Likert-type scale ranging from 1 = "never" to 5 = "always" (M = 2.00, SD = 0.50, Cronbach's α = .68). We included trait loneliness as a control variable for our analyses on social media use and affective well-being, and social media use, and momentary loneliness.
Social Well-Being and Self-Esteem: For the assessment of well-being, we used items of the KINDL-R questionnaire for adolescents aged 14-17 years (Ravens-Sieberer & Bullinger, 1998). Participants answered these questions on a 5-point Likert-type scale ranging from 1 = "never" to 5 = "always". The questionnaire included questions to measure social well-being related to friends (three items, M = 3.83, SD = 0.65, Cronbach's α = .62), social well-being related to the family (three items, M = 4.31, SD = 0.74, Cronbach's α = .79), and self-esteem (three items, M = 3.47, SD = 1.05, Cronbach's α = .85). We included social well-being and self-esteem as covariates in our analyses explaining affective well-being and loneliness.
Sociodemographic Variables: We asked the participants to indicate their age in years and gender (i.e., boy, girl). Given that students were sampled from different schools (i.e., lower secondary school, high school), we also included the school as a covariate in our analyses.

Analysis
For descriptive purposes, we first ran zero-order correlations among the between-person and within-person variables from the momentary assessments (see Table 1). We then conducted multilevel linear regression analyses in R to test our hypotheses because our data represent longitudinal hierarchical data with several momentary observations (Level 1) nested within participants (Level 2, see Table 2). In a first step, we ran intercept-only models, also referred to as null-model, to determine the intra-class correlation (ICC) for our dependent variables and our social media use variables. To test for the direct associations between mobile social media use and affective well-being, we ran a multilevel model, where we defined affective well-being as the outcome variable. We estimated a nested random effects model with 95% bias-corrected bootstrap confidence intervals based on 5000 bootstrapped samples. We entered fixed effects for each type of social media use (i.e., messaging, posting, or browsing) which were person mean-centered. For controls, we added age, gender, school, social well-being with friends, social well-being with the family, self-esteem, and trait loneliness as fixed effects. We allowed the intercepts to vary (= Model 1). We repeated these steps with momentary loneliness as our dependent variable (= Model 2). Finally, we tested the two-way interactions of FoMO and different types of social media use on affective well-being. We then repeated this step with loneliness as the dependent variable. For this purpose, we calculated grand mean-centered FoMO before computing the interaction terms. All intercept terms were random, the slopes were fixed.

Descriptive Results
We assessed the intercorrelations of the variables assessed during the mobile experience sampling, which are displayed in Table 1. Values above the diagonal represent zero-order correlations of within-person variables (observation at level 1), and values below the diagonal represent zeroorder correlations of between-person variables (observations at level 2). The variables for social media use (i.e., messaging, posting, and browsing) correlate positively with each other on the within-person and the between-person level. On the within-person level, messaging and posting were positively correlated with well-being. On the between-person level, only the correlation between messaging and well-being was significant and positive. There were no significant correlations between different types of social media use and loneliness on the within-person level, but on the between-person level, messaging and loneliness were negatively correlated. Well-being and loneliness were negatively correlated to each other on both analytical levels.
The intercept-only models provided information on the ICC for affective well-being and loneliness. The ICC indicated that 38% of the variance of affective well-being and 43% of the variance of loneliness could be traced back to stable differences between participants. The ICC for messaging (.277), posting (.458), and browsing (.428) equally indicate that a considerable amount of the total variance can be explained by differences between participants. Together, these values justified a multilevel analysis by including level 1 predictors. Figure 2 depicts the key results of our study. Supporting H1, we found that messaging was related to higher levels of affective well-being (b = 0.09, SE = 0.04, p = .024, ß = 0.07, see Table 3). In line with H2, we also found a positive association between posting and affective well-being (b = 0.21, SE = 0.07, p = .005, ß = 0.07). Answering RQ1, we found no significant association between browsing and affective well-being (b = À0.03, SE = 0.05, p = .544, ß = À0.02). The random part of the model showed that differences between participants explain about 17% of the total variance of affective well-being.

Testing the Moderating Role of FoMO
Answering RQ3a, we found a significant moderation of FoMO on the relationship between browsing and affective well-being (b = 0.15, SE = 0.06, p = .011, ß = 0.08). Probing the moderation revealed that for individuals with high levels of FoMO (above 3.99 or higher), browsing was associated with higher levels of affective well-being. For those with low levels of FoMO (1.95 or below), browsing was negatively related to affective well-being. We found no significant moderation of FoMO on the relationship between the other two types of social media use and affective well-being (messaging: b = À0.09, SE = 0.05, p = .066, ß = À0.05; posting: b = 0.03, SE = 0.10, p = .743, ß = 0.00).
Looking at the moderation as asked in RQ3b, we found no evidence of moderation of FoMO between messaging and loneliness (b = 0.02, SE = 0.05, p = .641, ß = 0.01) nor between browsing and loneliness (b = 0.00, SE = 0.05, p = .985, ß = À0.06). We found a significant association for the relationship between posting and loneliness, b = À0.19, SE = 0.09, p = .034, ß = À0.06. Probing the significant association revealed for those with FoMO level of 3.06 or higher, posting was associated with reduced loneliness. For those with lower levels of FoMO, the association was not significant.

Discussion
Social media, such as WhatsApp or Instagram, have received increasing importance in youth's daily lives, resulting in being permanently connected to others via their smartphones (Vorderer et al., 2016). Researchers have pointed to both opportunities and risks offered by social media on young individuals' psychological well-being. The current study aimed at providing further empirical insights into this prosperous research field. Informed by media richness theory (Sheer & Chen, 2004) and social presence theory (Short et al., 1976), we investigated the association between youth's different types of social media use, affective well-being, and loneliness by using a mobile experience sampling approach. By differentiating between messaging, posting, and browsing, we aimed to shed further light on different types of social media use.
Our findings showed that messaging and posting were positively related to youth's momentary affective well-being-while controlling for self-esteem and social well-being from the background survey. In terms of media richness, both types can be considered as richer types of media use when compared to browsing as these types of use provide the opportunity for online selfdisclosure and feedback. Several mechanisms, such as positive feedback, perceived connectedness, social support, capitalization, and psychological authenticity (Luo & Hancock, 2020;Valkenburg et al., 2017), may explain the positive association between online-self disclosure and affective well-being. Partially corroborating earlier findings among adult samples, we found no evidence for a direct relationship between browsing and adolescents' affective well-being (Burke & Kraut, 2016;Weinstein, 2017). However, we found a moderated relationship of browsing and affective well-being. On the one hand, for those with high levels of FoMO, following the desire to check social media by browsing its content was associated with higher levels of affective wellbeing. On the other hand, for youth with low levels of FoMO, browsing was associated with reduced affective well-being. Our results might, therefore, complement previous findings by Beyens et al. (2020), where a positive association between browsing and affective well-being was found.
When looking at loneliness, we found that browsing resulted in higher levels of momentary loneliness. Finally, the relationship between posting and loneliness is moderated by Fear of Missing Out. Participants with high levels of FoMO reported lower levels of loneliness after having posted content on social media. These results are partially in line with previous findings (Chai et al., 2019). Drawing on SDT (Ryan & Deci, 2000), we argue that FoMO might tap into the basic need of relatedness. In line with this reasoning, posting can also be understood as a means of responding to perceived stress to remain popular online (Beyens et al., 2016;Rideout & Robb, 2018). By posting content on social media, young individuals might make sure to remain up-todate, and they show that they equally have experiences to share online. Possible feedback via likes and comments from their peers might explain why they feel less lonely.
Before discussing the theoretical and practical implications of our findings, it is important to note that, in line with previous research, the effect sizes in our study were considerably small (Orben, 2020). Given that these statistics have been observed on an aggregate level, there have been recent efforts to disentangle underlying susceptibility factors on the individual level to study truly person-specific effects (Beyens et al., 2020;Pouwels et al., 2021;Valkenburg et al., 2021).
Our study showed that the associations between social media use and psychological well-being depend on the type of use. Specifically, social media use that involves elements of interaction or self-presentation (i.e., messaging and posting) appears to be beneficial for young individuals' affective well-being. Our findings regarding browsing, that has been traditionally understood as a passive type of use, are less straightforward. This type of content consumption, might result in detrimental outcomes, as in our case, increased momentary loneliness. For those with high levels of FoMO, browsing might be a positive experience.
These findings can be partially explained in line with media richness theory (Daft & Lengel, 1986;Sheer & Chen, 2004) and social presence theory (Short et al., 1976), suggesting that messaging and posting present communicative activities that allow for immediate reactions (i.e., receiving messages, comments, and likes) and thus convey feelings of social proximity. Browsing, however, does not involve feedback or reactions from others and rather elicits upward social comparison and envy. By and large, these findings corroborate meta-analytical findings on digital media use and psychological well-being . However, our findings also stress the importance of individual predispositions that might call long believed relationships of social media use and well-being into question-as, for example, Fear of Missing Out.
Hence, based on the direct associations of this study alone, it would appear plausible to argue that active private and active public use represents a beneficial type of social media use, while passive public use represents a detrimental type of social media use. However, in light of the moderated findings, this conclusion might be too simplistic. Specifically, with regards to so-called passive social media use, researchers begun to question the passive-active paradigm (Beyens et al., 2020;Ellison et al., 2020;Kross et al., 2021;Valkenburg et al., 2022). While browsing, intraindividual motivations and processes (e.g., inspiration, malicious and benign envy) might lead to positive, negative, or no effects among social media users (Meier & Schäfer, 2018;Nabi & Keblusek, 2014;Valkenburg et al., 2022). Furthermore, it has been argued that active practices of social media use need clearer operationalizations, non-clicking behavior on social media, or nonuse of mobile social media altogether result from conscious and thoughtful engagement (e.g., Ellison et al., 2020;Karsay & Vandenbosch, 2021;Matthes et al., 2020).
Our findings add to the public and scholarly debate about social media use and well-being among young individuals by highlighting the possible beneficial and adverse effects of different types of social media use. Instead of concluding whether social media are "good" or "bad" and, therefore, feeding the misconception of media effects, we would like to acknowledge the complexity of this field of research as outlined above. Apart from the type of social media use (e.g., messaging, posting, and browsing) and individual susceptibility factors (e.g., FoMO), also the amount of use, the balance next to non-media related activities, the wider social and cultural context needs to be taken into account in order to understand such media effects (Orben, 2021). Instead of overgeneralizing the study results, they should be understood as elements of a broader digital media diet. Furthermore, future research should also consider the actual content to explain possible media effects, for example, by linking exposure to a certain type of media content with various indicators of well-being or ill-being.

Limitations and Future Research
This study is not without limitations. We address some of them below and also provide recommendations for future research. First, the literature on mobile experience sampling recommends prompting participants at random points in time to capture a random sample of measurements. However, we prompted our participants to minimize the students' distraction from their school tasks prearranged times. Future research could consider sending alerts randomly during the day, which might provide even more detailed insights into the postulated relationships.
Second, although the mobile experience sampling design represents an up-to-date research methodology to smartphone media use and its influences, participants may experience surveytaking fatigue due to repetitiveness. To mitigate this possible risk, we provided easy-to-answer and straightforward questions, and we kept the length of the background survey and of the momentary surveys to a minimum. Thus, in line with previous mobile experience sampling research (MES; Beyens et al., 2020;Kross et al., 2013;Reissmann et al., 2018), we used single items to assess loneliness and affective well-being among youth. We recommend future studies using short scales, which have been validated in mobile experience sampling research (van Roekel et al., 2015). With regard to the shortcomings in the reliability of our measures, we also recommend using the full scale for FoMO and other measures for social well-being.
Finally, the current study investigates three types of social media use: active private (messaging), active public (posting), and passive public (browsing). Therefore, one type of use, passive private use (e.g., receiving messages), has not been addressed in our study.

Conclusion
Messaging, posting, and browsing represent daily media practices among youth. Before school, in school, and in the evening, middle and late adolescents in our sample used their smartphones to send messages, present themselves online, or browse through profiles of other social media users. We found that messaging and posting were positively associated with affective well-being, while browsing was associated with higher levels of momentary loneliness.
With our findings at hand, we provided several contributions to the field of research. First, we implemented a mobile experience sampling design that has only been used hesitantly in communication research. Instead of asking participants to indicate their general social media use, we asked about youth's mobile social media use within the last hour, several times per day. Thus, it is likely that we produced more reliable data than data stemming from traditional survey designs. Second, our study further emphasizes the need to include different types of mobile social media use (i.e., messaging, posting, and browsing) and moderator variables in this field of research since we found empirical evidence for moderated relationships. Including these differentiated measures allowed us to identify important boundary conditions, further explaining the relationship between mobile social media use and loneliness. Finally, our study highlights the need to investigate within-person and between-person differences because we showed that, especially when studying mobile social media use, methods and measures that account for daily fluctuations in psychological well-being are needed. Traditional cross-sectional or panel survey designs included traitlike measures. However, these measures are less suitable for detecting daily change in mood and emotions influenced by mobile social media use.

Research Transparency Statement
The authors are willing to share their data, analytics methods, and study materials with other researchers. The material will be available upon request.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Bundesministerium für Bildung, Wissenschaft und Forschung (SMARTPhone -SPA 06/109) and Fonds Wetenschappelijk Onderzoek.