It was expected that SME and the different SMA (passive, active and interaction) would be predictive of wellbeing, mood and sleep quality. The critical role of comparative information-processing online, would indicate that SCO would be predictive, as would optimism, given the evidence of its efficacy in relation to in-person and online interactions. For the results with sleep quality (Table 2), only optimism featured – increases in optimism were negatively associated with poor sleep quality. This result is consistent with Hernandez et al., (2014) and Lemola et al., (2013).
Where wellbeing was the outcome measure (Table 3), both optimism and SCO were significant. Optimism explained a larger variance in wellbeing. This supports the Qi et al., (2012) meta-analysis on the beneficial role of optimism on health and, consistent with Liu et al., (2017) and Festinger’s conception of social comparison theory, a negative relationship was observed between SCO and wellbeing. The result indicates that the type of comparisons respondents made had an adverse impact on wellbeing. Given upward comparisons online are more frequent (Twenge, et al., 2018), this finding suggests it was this type of comparison that accounted for the harmful effect of SCO on wellbeing (Vogel et al., 2015; Tandoc, Ferrucci & Duffy, 2015).
For the regression with depression (Table 5), optimism was the only significant predictor and it negatively predicted depression scores. With just one predictor, the variance explained was large, suggesting it was an important coping ingredient and supports Qi et al., (2012).
For the regression with anxiety (Table 4), optimism was the strongest predictor – increases in optimism predicted low anxiety. SCO was predictive and, as with wellbeing, it had an adverse effect on anxiety. This supports Liu et al., (2017). The only evidence for social media influencing any DV was with passive Instagram predicting anxiety. This is consistent with Yang’s (2016) finding in relation to passive Instagram use and loneliness.
This model reported evidence of optimism as a moderator between SCO on anxiety (figure 1). A positive relationship was observed between SCO and anxiety, irrespective of scores on optimism. However, the results revealed that those high on optimism scored lower on anxiety. This suggests that optimism acted as a buffer against the adverse effects of SCO on anxiety. Liu et al., (2017) found the same pattern with depression. The results suggest that optimistic thinking allows the individual to interpret comparative information in a way that reduces its potentially negative effects. The value of comparative information-processing online gives SCO particular potency. This is indicated by its positive association with anxiety and supports Tosun and Kasdarma (2020). The value of optimism to act as a buffer suggests there is merit in the cognitive strategies, such as defensive optimism, catastrophizing fantasies and reframing, more frequently used by optimists (Gibbons, 2022a, 2022b).
While the current study sought to explore the relationship between SCO and SMU, the items on the SCO scale are context free (Gibbons & Buunk,1999). They make no reference to the comparisons made online as distinguishable from in-person comparisons. The items, in fact, appear to be designed for in-person comparisons e.g. ‘I often like to talk with others about mutual opinions and experiences’ [item 7]. While most can be interpreted as relevant for virtual as well as in-person comparisons, its lack of specificity to social media may question its validity for this context and the results may reflect the importance of SCO in relation to anxiety for both in-person and virtual comparisons.
Mediation analysis
Positive relationships were found between the frequency of passive Instagram use and anxiety and SME and anxiety, and similarly for wellbeing in place of anxiety (Figures 2 & 3, Table 6). However, the relationship was no longer significant when SCO was added. This suggests that SCO was a mediator - those scoring high on SCO were more likely to engage online and, independently, those scoring high on SCO, more frequently scored high on anxiety. This implies that the attempts to draw comparisons with the lives of others was, for those scoring high on SCO, more disruptive to anxiety and wellbeing. Twenge et al., (2018) drew the same conclusion. Users that drew upward comparisons were more likely to interpret their life as less worthy and interesting. Twenge et al., (2018) argued that while users know that content shared on social media is frequently akin to a ‘highlight reel’ - of the best in others’ lives, they struggle to separate their emotional response from this cognitive understanding.
Twenge et al., (2018) sampled adolescents and Erikson (1994), in his psycho-social theory of development, argued that striving for identity achievement is critical during adolescence. One’s sense of identity is formed through the interactions and responses of others, first with family, then friends and peer groups. As the self expands, the reactions of one’s peer group become increasingly important. This may explain why social comparison with peers and peer pressure are highest during adolescence (Vogel et al., 2014). The sample in this study was broader than just adolescents (age range 18-79 years). It is possible that this finding suggests the tendency Twenge et al., (2018) demonstrated in adolescents, remains important in older cohorts. Erikson (1994) argued development was not confined to the family or to childhood, but to relationships outside the family and those formed across the lifespan. Using social media, not just daily but frequently throughout the day, is ubiquitous. It offers an opportunity to connect with and see the lives of others (Yang, 2016) and while the lives of those you are viewing may have a particular relevance during adolescence, they can remain important, for different reasons, across the lifespan (Erikson, 1994). This may offer an explanation for the evidence on SCO mediation found here.
SME and passive instagram use, a particular type of SMA, featured in the direct path with anxiety and with wellbeing (figure 3, Table 6) but these were no longer significant when SCO was added. Both passive instagram and SME positively correlated with SCO. Passive instagram describes users passively browsing newsfeeds and profiles without leaving comments or engaging (Yang, 2016). If one is motivated to draw superficial comparisons rather than to understand and connect with others in a meaningful way, then passively scrolling is a logical choice. The most likely outcome from this is a judgement and emotional response to what one’s (unedited) life is like compared to the edited highlights of others (Twenge et al., 2018). It is a type of comparison likely to lead to disappointment (Yang, 2016) and, in this study, to increased anxiety and lower wellbeing. This is consistent with the adverse effect of upward social comparisons (Tandoc, Ferrucci & Duffy, 2015; Twenge, et al., 2018).
Optimism featured as a mediator between SME and wellbeing (figure 4, Table 6). SME was negatively associated with wellbeing but was no longer significant when optimism was added. Optimism negatively related to SME and positively with wellbeing. This suggests that the amount of SME is significant, less is more beneficial and is associated with higher optimism scores and it implies that the nature of that engagement is important. This is consistent with the different impacts that passive, active and interactive engagement have on wellbeing (Smith, 2000; Yang, 2016). While the specifics of this were not measured, it is likely that those more optimistic brought the same thinking patterns used in their in-person life to their online life (Gibbons, 2022a, 2022b; Liu et al., 2017). For example, they may look to form more meaningful connections, offer more positive comments, make more compliments and be biased to attend more to positive messages over toxic ones.
Limitations and improvements
As the sample was convenience and volunteer based and given the non-response rate for some of the measures, validity could be an issue. Order effects and state congruence recall may also have affected the completion rate and response validity. The use of a single attention-testing item (e.g., please choose a specified response below) in each scale and excluding respondents who did not accurately answer these attention-testing items is likely to increase validity and could form part of the screening plan.
Optimism featured in all the regression models – it is beneficial for sleep, mood and wellbeing. It featured as a moderator with SCO and as a mediator between SME and wellbeing. The results suggest that optimistic strategies appear to be beneficial, not just in face-to-face interactions but in online ones (NB the SCO items were context free). It would be fruitful to explore this further i.e., how those high in optimism engage on SMU compared to those low in optimism. The mediation result reported an inverse relationship – those high in optimism engaged less. However, this study leaves unanswered the nature of that engagement and cognitive strategies used. Exploring this further could inform the current user-guides on healthy tips for SMU. These currently focus on encouraging less time on devices, checking the credibility of sources and encouraging click restraint (Bartolomeo, 2020) but do not consider such cognitive strategies.
SCO was associated with adverse anxiety and wellbeing. On SMU, increased social comparison tendencies were harmful. Passive engagement, such as scrolling feeds and profiles, is more frequently engaged by those high in SCO. It is more frequently associated with upward comparisons and adverse health (Twenge et al., 2018; Vogel, Rose & Roberts, 2014). Optimists engage less frequently on SMU and adopting the strategies employed by optimists, irrespective of one’s level of dispositional optimism, is likely to be advantageous to mood, wellbeing and even sleep. It influences how one processes comparative information in in-person interactions and on SMU.
Fredrickson and Losada (2005) took a positive psychology approach to nurture more productive business meetings in the corporate sector. They found a 3:1 ratio of positive to negative comments/statements was the optimal ratio associated with greater economic success. There may be an optimal ratio to time on devices, to time scrolling, posting, commenting etc. The analysis in this study points to the merit of exploring this and the types of cognitive strategies associated with healthy SMU.