Correlates of social media fatigue and academic performance decrement: A large cross-sectional study

2.1 Social media fatigue

A review of the prior body of research on social media fatigue indicates that this phenomenon has gained scholarly attention only recently (Table I). From a platform perspective, most studies have concentrated on different aspects associated with fatigue and stress induced through the most notable online social media platforms, such as Facebook (Ravindran et al., 2014; Bright et al., 2015; Cramer et al., 2016; Dhir et al., 2019; Lee et al., 2019), Qzone (Zhang et al., 2016), and social media platforms in general (Logan et al., 2018; Lee et al., 2016; Pontes, 2017; Salo et al., 2017). In comparison, a handful number of studies have been conducted in the context of mobile instant messaging (MIM) apps (Shin and Shin, 2016; Sun et al., 2017; Xiao and Mou, 2019)

Most of these studies have been carried out with participants in developed economies, including the USA (Bright et al., 2015; Cramer et al., 2016), South Korea (Lee et al., 2016; Shin and Shin, 2016; Lim and Choi, 2017), Portugal (Pontes, 2017) and Finland (Salo et al., 2017). More recently, studies from emerging economies such as China, India and Pakistan have also started to emerge (Zhang et al., 2016; Luqman et al., 2017; Dhir et al., 2018, 2019). Our review clearly indicates that more empirical studies on social media fatigue are needed in the context of emerging economies such as India. This will augment our current understanding of social media users, which, until now, has predominately focused on users from the USA and developed countries. Moreover, the cultural diversity lens will further offer novel insights and comparisons about the actual usage patterns and preferences of social media users in different regions of the globe.

Prior literature suggests that scholars have used a number of theoretical frameworks to understand social media fatigue. Most of these frameworks originated from psychology, for example, the stressor-strain-outcome (SSO), stress coping model, stimulus-organism-response (SOR) model and transaction theory of stress (Shin and Shin, 2016; Zhang et al., 2016; Luqman et al., 2017; Salo et al., 2017; Dhir et al., 2019). Studies on social media fatigue have mostly examined social media's correlations with different social media use behaviors, such as compulsive use, fear of missing out, privacy concerns and information, communication, system and social overloads (Bright et al., 2015; Beyens et al., 2016; Salo et al., 2017).

Previous research on the topic has also determined the potential repercussions that may arise because of social media fatigue; this includes an investigation into the possible consequences or outcomes, such as suspended or reduced activities (Ravindran et al., 2014; Shin and Shin, 2016; Zhang et al., 2016; Luqman et al., 2017). In addition to this, many studies have pointed toward the negative psychological and physiological outcomes of experiencing social media fatigue, such as emotional exhaustion (Lim and Choi, 2017), depression and anxiety (Pontes, 2017; Dhir et al., 2018), sleep deprivation, concentration issues and problems in social relationship management (Salo et al., 2017).

Building on the current literature, the present study investigates whether different measures related to social media use behavior and academic well-being (i.e. academic performance decrement) correlate with social media fatigue. Unlike most prior studies, the current study considers fatigue by looking at an MIM app (i.e. WhatsApp) – one of the leading and most popular social media platforms in India and other countries. Despite being a leading MIM platform, WhatsApp is positioned under the umbrella of social media that accommodate Web 2.0 applications, where users create and exchange their own content (Kaplan and Haenlein, 2010). WhatsApp or MIM in general is different from other prominent social media platforms, including Facebook and Instagram. The main core characteristics of WhatsApp distinguishes it from other prominent platforms. These are: first, WhatsApp supports day-to-day communication in synchronous mode with others (individual and group) in the form of text, voice and video. The interactions on WhatsApp are not intended for the public at large because they are often limited to close contacts or specific social groups. Second, WhatsApp enables its users to share multiple content formats and offer a free alternative to network calling, which sets it apart from other forms of social media. Our observations are consistent with the prior literature on new media, which suggests that the users of different social media use these platforms differently because of underlying affordances and associated gratifications (Miller et al., 2016).

Considering the above premise, the current study expands on and complements the existing literature on the possible antecedents and outcomes of social media fatigue by using lesser known measures in the context of young adult WhatsApp users.

2.2 Online privacy concerns and fatigue

User privacy has been one of the most actively debated topics in the context of social media platforms. Social media platforms are specifically designed to lure users to become active content generators. Maintaining a personal profile and sharing content within the network and publicly exposes people to a number of privacy and security risks, such as identity theft, profile hacking, cyberbullying and online surveillance (Krasnova et al., 2009; O'Keeffe and Clarke-Pearson, 2011). These risks from network externalities translate into user fears, which are often termed privacy concerns. In the present study's context, privacy concerns refer to the users' degree of anxiety about the possibility that their personal, confidential or sensitive information may be monitored, exposed or violated by others within or outside the network.

As social media proliferates and becomes an integral part of the daily lives of many online users, issues associated with privacy become ever more significant. Recent developments, such as the Cambridge Analytica controversy, is one example that raises alarm regarding the privacy of user data (Kuchler, 2018). Recent research indicates that privacy concerns of social media users negatively influence their behaviors, activity levels and trust with these systems (Taddicken, 2014; Malik et al., 2016a, b). Furthermore, being too accessible or getting too much information through these platforms also contributes to privacy concerns among users (Bright et al., 2015). Because of active mentions of privacy issues in the media, stories of privacy breaches, uploaded private and sensitive data and somewhat complex and confusing privacy settings, social media users feel pressured, which can potentially contribute to fatigue.

There are two possible reasons for the possible association between online privacy concerns and social media fatigue. First, when a task, event or activity is executed, then it results in the utilization of a person's cognitive space, which, by its very nature, is limited (Boksem and Tops, 2008). Similarly, when a social media user is too concerned about his or her online privacy, his or her cognitive space is also affected in a way that could lead to social media fatigue. Second, social media users with high privacy concerns are likely to become anxious about their social media use, and this translates into social media fatigue (Bright et al., 2015). However, compared with this, some scholars also believe that social media users with low online privacy concerns are likely to become too accessible or might receive too much content and information from other people (Bright et al., 2015). This is also consistent with the observation of Schwartz (1968, p. 741) that beyond a given threshold, keeping social contacts becomes irritating and thus difficult. Because of this, it is likely that social media users with low privacy concerns might experience social media fatigue. Therefore, we put forward the following hypothesis:

2.3 Online social comparison and fatigue

Social media platforms are excellent channels for self-presentation because they facilitate users in generating an image of themselves; the image created on these platforms influences how people within their network perceive and value a particular user (Goffman, 1959). An extended review of social psychology literature shows that individuals comparing themselves with those who they think are superior (or better-off) may result in negative emotions, including envy or depression. Furthermore, these negative emotions can have a negative impact on health and psychological well-being (Buunk and Ybema, 2003).

Social comparisons occur in various offline contexts, i.e. daily life, but social media platforms simplify and expedite the whole process (Cramer et al., 2016). For instance, Cramer et al. (2016) revealed that almost 70 percent of their study participants compared themselves with their Facebook friends. Considering this motive, users tend to post socially desirable and glossy portrayals that aim to optimize a positive self-image and impressions among friends and followers (Mehdizadeh, 2010; Cramer et al., 2016). In doing so, social media users tend to share crafted and selective information about their lives that might not reflect reality. On social media, many users do not personally know or have not met most of their connections, and hence, users tend to form an image of each connection based on shared content (e.g. status updates, pictures, likes and comments). This shared content generally depicts people's lives based on more positive characteristics, such as success, family, wealth and appearance, which might not be true (Qiu et al., 2012). In line with the social comparison theory, these positive life experiences, events and feelings lead other users to be more prone to self-comparisons because they tend to relate that information to themselves and compare their lives to other people's lives (Festinger, 1954).

This exposure to content depicting highly favorable life experiences and gauging one's abilities, performance and life in general can have negative psychological outcomes that negatively impact one's well-being. Because users have been increasingly using their social media for social comparisons, there is a growing interest in the resulting psychological outcomes. The literature has reported a number of psychological and behavioral reactions, such as envious feelings, emotional distress, jealousy and low self-confidence, as a result of comparing oneself with others on social media platforms (Feinstein et al., 2013). There is also significant empirical support to suggest that engaging in social comparisons on these platforms (particularly Facebook and Instagram) poses negative psychological outcomes that are likely to affect users' psychological well-being and mental health, leading to poor life satisfaction, mood swings and depression (Vogel et al., 2015; Kross et al., 2013). Based on the above premise, we argue that carrying out a social comparison on MIM will result in fatigue. Therefore, we hypothesize the following:

2.4 Online self-disclosure and fatigue

Sharing information about oneself is one of the key motivations for why individuals use different social media platforms (Qiu et al., 2012; Bazarova and Choi, 2014). Compared with other forms of online channels (e.g. content communities, blogs or collaborative projects), the level of self-disclosure is far higher on social media platforms (Kaplan and Haenlein, 2010). Most of these platforms require users to create a personal profile by adding a display photo and personal information (e.g. name, gender, location and interests).

Similar to communication in offline environments, self-disclosure or sharing personal information is somewhat necessary to effectively use the various features of these platforms and become involved in self-expression, relationship initiation and relationship maintenance (Altman and Taylor, 1973). As the number of social media users grows and the amount of time these users spend on these platforms also increases, many of them are inclined to unmask and share their personal stories, thoughts, feelings and experiences in public and privately. Many users engage in self-unmasking on these platforms with the intention of forming and bonding intimate, meaningful relationships (Bazarova and Choi, 2014). To achieve these objectives, they disclose personal, sensitive and confidential content without realizing the inherent and somewhat established notion of losing control of their disclosures (Bazarova and Choi, 2014).

A fair amount of literature has highlighted that because of disclosure of sensitive personal information through online self-disclosure, online users become prone to ramifications such as cyberbullying, unwanted sexual solicitation, sexting, online harassment and internet addiction (Krasnova et al., 2009; O'Keeffe and Clarke-Pearson, 2011). For young people, particularly adolescents and young adults, the likelihood of such events in the case of young people is quite high (O'Keeffe and Clarke-Pearson, 2011). Scholars have argued that although these vulnerable groups are being educated about the potential threats present on social media platforms (Vanderhoven, 2014), these groups are tempted to disclose sensitive information because of peer influence and expected rewards (Sheldon and Pecchioni, 2014). In addition, a growing body of literature has linked online self-disclosure with negative psychological well-being, such as anxiety, fatigue and mental stress (Boksem and Tops, 2008; Rains et al., 2016). Based on these findings, it is likely that self-disclosure will exert strain on MIM users in the form of social media fatigue. We hypothesize the following:

2.5 Intensity of MIM use and fatigue

In the context of social media, across the glove, there is an upward trend in the daily time spent by users. A recent report published by Statista (2019) highlighted that, on average, Web users spend over 2 h (136 min) every day on different social media channels. The excessive use of technology among different cohorts of society has been an active research topic. Studies from a number of disciplines have indicated that high exposure to contemporary gadgets and technologies (McDaniel and Coyne, 2016), mobile phones (Eyvazlou et al., 2016) and social media (Rosen et al., 2013a, b) burdens user behaviors and has negative consequences on psychosocial and physiological well-being. For instance, the study by Thomée et al., (2010) looking at young Swedish adults (ranging from 20 to 28 years old) reported an association of high ICT usage with various mental symptoms, including frustration, role conflicts, time pressure, mental overload, neglect of personal needs and other activities, social isolation and guilt feelings. In another study with adults ranging from 18 to 65 years old (mean = 30.74), the researchers observed a number of clinical symptoms among the heavy Facebook users, including major depression and schizoid disorders (Rosen et al., 2013a, b). Furthermore, a number of recent studies have also reported the impact of social media intensity, including negative psychological outcomes such as loneliness, low self-esteem, depression and exhaustion (Karapanos et al., 2016; Cao et al., 2018; Dhir et al., 2018). Therefore, it is plausible to hypothesize that the intensity of MIM use is positively associated with social media fatigue. We propose the following hypothesis:

2.6 Fear of missing out (FoMO) and fatigue

The fear of missing out (FoMO) is a prevalent unease about missing information, rewards or gratifying experiences that other members of one's network might be enjoying (Przybylski et al., 2013; Elhai et al., 2016). This personality trait connects with the desire to constantly connect with and follow others on different social networks. Prior literature on FoMO has suggested that the desire to persistently stay connected online to monitor or communicate with others is likely to result in the excessive use of mobile phones (Elhai et al., 2016; Cao et al., 2018), the internet in general (Kandell, 1998), online gaming (King et al., 2017) and social media platforms, particularly Facebook (Przybylski et al., 2013; Alt, 2015; Beyens et al., 2016; Dhir et al., 2019). FoMO on social media can be linked to the impulsive desire or urge to connect on these platforms and engage in activities, for example, chat with others, share or view updates, read or respond to conversations/comments or play social games. On social media platforms, FoMO is usually apparent when the user impulsively wants to connect in inappropriate situations, for example, right after waking up, during lectures or meetings or while driving (Przybylski et al., 2013; Alt, 2015; Hetz et al., 2015).

Like other forms of technology-related anxieties and problematic use, scholars have attributed FoMO to negative physiological and psychological well-being, such as reduced life satisfaction, low self-esteem, inadequacy, high alcohol consumption, depression and physical issues (Przybylski et al., 2013; Elhai et al., 2016). Moreover, individuals with the feeling of FoMO are more likely to have negative emotional feelings, social exclusion and social envy (Hetz et al., 2015). Although FoMO is prevalent in all cohorts of online users, recent studies have reported a higher level of FoMO among adolescents and young adults, i.e. those attending college. For instance, the study carried out by Beyens et al. (2016) reported that adolescents with a higher need for belonging and popularity on Facebook experience higher levels of FoMO, which results in intense use of the platform. Furthermore, Facebook users with high levels of FoMO experienced more stress (Beyens et al., 2016). Based on the above premise, we hypothesize the following:

2.7 Academic performance decrement and fatigue

A number of studies have investigated the adverse implications of technology in the workplace and its effects on learning and within social contexts (Ricci et al., 2007; Rosen et al., 2013a, b). These studies found strong evidence of a negative correlation between excessive use of technology and induced stress or fatigue with human performance and output, for example, in gaming (Jackson et al., 2011) and mobile phones (Lepp et al., 2015).

Technologies aimed at gratifying hedonic needs, for example, social media platforms, have very high cognitive absorbing attributes that usually increase users' involvement and dependency (Cao et al., 2018). Spending excessive time on these platforms within and outside the classroom can limit the time spent studying and concentration. Mental preoccupation with these platforms during academic activities will result in an upsurge in psychological stress that ultimately can lead to lower actual performance (Ayyagar et al., 2011). For instance, Kirschner and Karpinski (2010) carried out a study with 218 Facebook users, and they observed a lower self-reported grade point average (GPA) among heavy social media users. Likewise, other studies with US college students found that increased interactions and usage of Facebook contributed to lower actual GPAs (Junco, 2015; Rosen et al., 2013a, b). Some studies have also reported that lower GPA scores are the outcome of excessive mobile phone usage, for example, texting and calling (Hong et al., 2012).

The aforementioned studies in educational settings have measured academic performance by using GPA in a particular subject or GPA for the overall year. Compared with this, the present study has used a more holistic measure: academic performance decrement, first coined by Kubey et al. (2001). Academic performance decrement measures the negative impact of social media use on the overall academic performance of a student. For example, it measures whether social media use hurts or disturbs college work and sleep, resulting in exhaustion and a lower GPA. Based on the extended social media literature, we hypothesize that social media fatigue is positively correlated with academic performance decrement because of social media use, as follows:

3.1 Research model

To understand and determine the different antecedents and outcomes associated with social media fatigue, scholars have used a number of theoretical frameworks predominantly originating from psychology. For instance, the SOR framework has been used to examine whether excessive Facebook use can lead to discontinuance usage intentions (Luqman et al., 2017). Salo et al. (2017) applied the transaction theory of stress to investigate the link between various social networking sites (SNS) stressors and individual/social strains among Finnish social media users. In another study with users of a popular mobile messenger in South Korea – Kakao Talk – researchers used a stress coping model to understand the impact of overload with shunning behaviors (Shin and Shin, 2016). Moreover, push-pull-mooring, the person-environment fit model, the limited capacity model and the SSO have all been used to investigate the antecedents and outcomes of various social media platforms (Sun et al., 2017; Lee et al., 2016; Bright et al., 2015; Kim et al., 2019; Cao et al., 2018; Yu et al., 2019).

To empirically explain the fatigue process associated with MIM use, the research model of the current study uses the SSO framework. We specifically opted for the SSO framework because it perfectly aligns with our study objective, i.e. correlates social media fatigue in terms of its antecedents and outcomes. Koeske and Koeske (1993) originally proposed the SSO framework, and it helps to explain how behavioral outcomes result from different stressors. The SSO illustrates a comprehensive account of the interplay of three major components: stressor, strain and the resulting outcomes (Figure 1). The framework postulates that humans tend to alter their behaviors (outcomes) to minimize the aftermath of detrimental reactions (strains). These strains usually emerge if there is an imbalance between the situation of a person and a requirement of the environment (stressor; Koeske and Koeske, 1993). Because of its origins in occupational health psychology, the model has also been widely used to study stress and fatigue induced through technology use and its outcomes among users (Ayyagari et al., 2011).

In the current study's research model, the stressors are different stimuli represented by social media use behavior, namely, online privacy concerns, social comparisons, self-disclosure, the intensity of MIM use and FoMO. Based on the SSO model, these stressors can trigger various emotional states that are referred to as strains (i.e. social media fatigue). In the final stage of the process, strains could potentially lead to a number of psychological and physiological outcomes, for example, depression, anxiety (Pontes, 2017; Dhir et al., 2018), behaviors such as shunning or discontinuance (Ravindran et al., 2014; Luqman et al., 2017; Salo et al., 2017; Sun et al., 2017) and a downturn in performance (Lepp et al., 2015; Zhang et al., 2016). Compared with this body of literature, the present study has considered academic performance decrement as a possible consequence of social media fatigue.

3.2 Study participants

The present study consists of two cross-sectional self-reports organized with young adult students attending a large public university in northwestern India. The target population was 17–27-year-old young adults who actively use WhatsApp, a popular MIM application in India. We conducted Study A with 1,398 MIM users (19–27 years old) attending a full-time master's program at the university. The mean age of the study participants of Study A was 22.81 (SD = 4.13), where 57.4 percent (n = 803) were females. We organized Study B with 472 MIM users (18–23 years old) who were only attending a distance education course and were not full-time students at the university. The average age of the participants was 20.30 (SD = 3.13), where 63.1 percent (n = 298) were females. The participants of Study B were different from those participating in Study A because they were not full-time students and instead had full-time employment and were taking part in distance education for career-oriented excellence.

The process for self-reported data collection was as follows: we advertised the proposed study among the target population via different university departments, teachers and advertisement on notice boards. We informed the target user group about the different research objectives and questions, anticipated benefits and related research process. Afterward, we invited the interested students to participate in both studies with the assistance of the respective departments. We organized the survey answering sessions in a classroom environment and internal faculty members monitored the sessions. Before the survey answering exercise, we again informed the study participants about the research process, aims and objectives and anticipated benefits. The study participation was voluntary, meaning that the students could withdraw at any time. Participation was confidential and anonymous; we collected no personal information, except age and gender. We obtained informed oral consent before allowing the students to participate.

3.3 Data analysis

The current study examined the goodness of the model fit through different indices, namely chi-square/degree of freedom (X²/df), comparative fit index (CFI), Tucker–Lewis index (TLI) and root mean square error of approximation (RMSEA; Kline, 2015). The recommended values for a good model fit include X²/df < 3.0, CFI and TLI ≥ 0.92 and RMSEA ≤ 0.08 (Wang and Wang, 2019). A good model fit shows a close fit between the empirical data and the theoretical research model. We followed a two-step procedure recommended by Anderson and Gerbing (1988). First, we examined the measurement model using a confirmatory factor analysis (CFA) by accessing model fit indices and different forms of instrument validity and reliability. In the second step, we estimated the structural model using structural equation modeling (SEM) to answer the different research hypotheses.

4.1 Measurement model

Model A, based on Study A (X²/df = 4.54, CFI = 0.94, TLI = 0.93 and RMSEA = 0.05), and Model B, based on Study B (X²/df = 2.44, CFI = 0.94, TLI = 0.92 and RMSEA = 0.06), revealed good model fit (Hu and Bentler, 1999; see Table II).

4.2 Validity and reliability

We examined convergent and discriminant validity and construct reliability in the context of the study measures used in both the research models.

4.3 Structural model

The structural model for both Study A (X²/df = 5.42, CFI = 0.93, TLI = 0.91, RMSEA = 0.06) and Study B (X²/df = 2.63, CFI = 0.92, TLI = 0.91, RMSEA = 0.06) revealed good model fit. The percentage variance explained in MIM fatigue for study A was 59.8 percent, and for Study B, it was 70.2 percent. Similarly, the percentage variance explained in academic performance decrement due to MIM use for Study A was 30.9 percent and for Study B was 35.1 percent (Figures 2 and 3). The structural model suggests the intensity of MIM use was a significant predictor of social media fatigue for Study A (β = 0.61, p < 0.001) and Study B (β = 0.62, p < 0.001). Other significant predictors were social comparison (Study A (β = 0.22, p < 0.001); Study B (β = 0.18, p < 0.001)) and self-disclosure (Study A (β = 0.16, p < 0.001); Study B (β = 0.13, p < 0.05)). Compared with this, online privacy concerns and FoMO did not share any relationship with social media fatigue for both Studies A and B. The results also suggest that social media fatigue was a strong predictor of decrement in academic performance across Study A (β = 0.56, p < 0.05) and Study B (β = 0.59, p < 0.05). The study findings support hypotheses H2, H3, H4 and H6 (Table V).

6.1 Theoretical implications

To begin with the theoretical implications, the present study has outlined five main contributions to the theory and literature. First, the most significant contribution of the current study is the focus on an MIM app, namely, WhatsApp. Although a predominant number of recent studies have investigated social media fatigue in the context of SNS such as Facebook, interest in other social media applications and tools is increasing, for example, Qzone (Zhang et al., 2016) and Kakaotalk (Shin and Shin, 2016; Lim and Choi, 2017). Consequently, the present study contributes to the limited literature on fatigue from MIM apps. Our study is also consistent with the recommendations of social media scholars who have strongly emphasized the need to opt for a granular approach by studying different social media platforms because their affordances and relevant implications vary substantially (Bazarova and Choi, 2014; Dhir et al., 2019). The present study could possibly serve as a potential source for comparing the antecedents and outcomes of fatigue because of different social media platforms.

Second, the current study expands the prior literature on social media fatigue by incorporating an SSO framework. The present study examined a number of stimuli (stressors), emotional states of stress (strains) and the performance of respondents (outcome). By providing a comprehensive view of the whole process, the current study complements the initial work on social media fatigue and technostress, which only has examined the relationship between stressors and strains (Bright et al., 2015; Beyens et al., 2016; Cramer et al., 2016; Lee et al., 2016). Furthermore, the current study contributes to the growing literature on the psychological well-being of online users.

Third, the current study extends our understanding of the impacts of the outcomes of social media fatigue. Research has mostly investigated the outcomes of experiencing social media fatigue by looking at behavioral measures or response strategies, for example, activity suspension, switching to other platforms or control strategies (Ravindran et al., 2014; Lim and Choi, 2017; Luqman et al., 2017; Sun et al., 2017), ignoring the impacts on users' performance. To bridge this gap, the current study specifically concentrated on performance outcome (i.e. academic performance decrement) as a result of social media-induced fatigue. The results from the current study revealed that fatigue induced through WhatsApp can significantly and negatively impact students' academic output and. The present study will likely motivate other scholars to better understand the other negative outcomes of social media fatigue, especially in terms of well-being.

Fourth, the current research extends the prior literature on the negative effects or the dark sides of unregulated and inappropriate social media use. As social media platforms become more prevalent among society, there is a strong desire and need to explore and comprehend the downsides of these platforms because most of the research so far has predominantly concentrated on the general usage, motivations and affordances of these platforms (Malik et al., 2016a, b; Khan, 2017; Mamonov and Benbunan-Fich, 2017; Capriotti and Ruesja, 2018). The findings from the current study illustrate the potential of using the literature from social psychology to expand our current outlook on social media dependence. The model in the current study expands the prior literature regarding the downsides of unregulated and inappropriate social media use. Furthermore, by investigating two different batches of study participants, the validity and reliability of the current study's findings are significantly improved.

Finally, despite the upsurge in internet access and smartphone ownership in countries such as India, Pakistan and Turkey (Poushter, 2016), there has been a limited understanding of user behaviors and attitudes in these regions. The current study augments the growing body of knowledge on the negative implications of excessive use of social media platforms in emerging economies (Dhir et al., 2018, 2019; Zhang et al., 2016; Luqman et al., 2017).

6.2 Practical implications

The findings of the current study present a number of implications for social media users, service companies and designers, as well as parents and teachers.

First, examining and connecting the relationships between the various determinants of social media fatigue in an educational context can offer insights into the post-adoption effects and experiences of WhatsApp on students. Being informed about the possible correlations of social media fatigue, users can effectively manage their social media usage. Based on the current study's findings, users should carefully assess their usage and interactions with different social media platforms. In both of the observed groups, the intensity of use has been the strongest predictor of fatigue, calling for users to proactively consider and opt for strategies to moderate their usage.

Second, educational institutions should hold regular workshops and counseling services to address the issues and concerns related to social media use and abuse. A strong responsibility lies with parents because they should monitor their children's online activities and promote the fair use of social media and other technologies. Recent studies have indicated that the younger generation is more active in using social media and spends a major chunk of their day on these platforms (Wallace, 2015). There is also a strong indication that the unregulated use of modern technologies among youth can potentially have adverse psychological and physiological outcomes, such as depression, anxiety and stress (O'Keeffe and Clarke-Pearson, 2011; Pontes, 2017; Dhir et al., 2018).

Third, the current study's findings also provide a number of implications relevant to practitioners and technology companies in general. Getting insights into the antecedents and consequences of fatigue through different platforms will support companies and relevant stakeholders to tailor their content and marketing strategies in a way that can limit switching, inactivity and permanent deletion by users experiencing fatigue. Because the results from our study reveal a strong negative impact of social media fatigue on students' academic performance, social media developers also need to take their social responsibility role seriously. Rather than focusing on the intensity of usage, companies should strongly focus on facilitating users so that they can efficiently complete their tasks on social media platforms. Offering features and tools (e.g. number of logins, screen time and auto sleep/away mode setup) that can help their users control and monitor their own usage will not only reduce the associated fatigue but will enhance the overall user experience and brand loyalty.

Finally, the insights from the current study can also help developers and service designers make informed design decisions for new and upcoming features, allowing them to address the identified factors causing fatigue among users. For instance, to minimize the unfavorable outcomes in the form of fatigue, social media platforms should carefully consider new features that do not force or entice self-disclosure behaviors.