Subgroup Analysis and Social Media Mental Health Impact: A Comprehensive Overview

The experiment revealed that the improvements in emotional state were not uniform across all participants. In the case of Facebook, the data indicated that users over the age of 35 experienced larger enhancements in their emotional state than did younger users (aged 18-34). Specifically, the subgroup analysis showed statistically significant differences between these age groups (p = 0.046), suggesting that older users may gain more benefit from reducing social media engagement. Additionally, findings indicated marginally larger effects for undecided voters and for individuals without a college degree, with p-values of 0.053 and 0.058 respectively. Importantly, these subgroup differences persisted even after the analysis controlled for baseline emotional state and baseline usage patterns, implying that the variations appear to be inherent to the subgroup characteristics rather than driven by initial conditions^[1].

The results differed between Facebook and Instagram. For Instagram, a standout finding was that the largest improvement in emotional well-being was observed among young female users, specifically women aged 18-24. This subgroup experienced an improvement of 0.111 standard deviations with a p-value of 0.002. In contrast, the effects among other Instagram user subgroups were less than half as large and statistically indistinguishable from zero. Such a pattern underlines that while overall deactivation of Instagram did improve emotional state, it did so most notably for young women. This observation is consistent with broader public concerns about Instagram’s impact on this particular demographic and reinforces the idea that the mental health effects of social media may be mediated by the distinct nature of content and interactions on each platform^[1].

The heterogeneity in the findings is critical for understanding the complex role social media plays in mental health. For Facebook, older users’ larger gains could be reflective of different usage habits or the type of interactions they have on the platform, which may be more conducive to stress or negative comparisons. Even though baseline usage and initial emotional state did not significantly alter the treatment effects, the results point to the possibility that user characteristics like age and educational attainment play a decisive role in how social media affects mental health. For Instagram, the pronounced benefit seen among young women suggests that the platform’s visual and social dynamics might contribute to adverse mental health outcomes for this group when used intensively. Thus, reducing exposure for these users through deactivation seems to yield a substantial improvement in emotional state. Additionally, even though higher political engagement showed somewhat larger point estimates for both platforms, these differences did not reach statistical significance, indicating that the deactivation effects are primarily driven by factors other than exposure to political content during election periods^[1].

In conclusion, the subgroup results demonstrate that while social media deactivation generally leads to improved emotional states, the benefits vary significantly across different demographic and behavioral groups. Older Facebook users exhibit notably larger improvements, while young women on Instagram gain the most from deactivation. These findings are important for informing mental health interventions and public policy, as they suggest that personalized approaches may be more effective when addressing the adverse mental health impacts associated with social media. The results underscore the need to consider demographic and usage patterns when evaluating the overall well-being effects of social media platforms^[1].