The role of attachment in shaping psychosocial processes has been studied extensively in the offline context, but little is known about parallel online processes. This study examined whether associations between attachment anxiety and responses to potentially negative partner behaviours (presented using vignette methodology) would be moderated as a function of whether hypothetical scenarios occur in online (face-to-face) or offline (social media) contexts. A total of 267 participants completed a measure of adult attachment (RAAS), and a relationship events questionnaire (R-REQ), in which participants rated their psychosocial responses to potentially distressing partner behaviours featured in online and offline scenarios (median age: 21 years; 81% female). Positive associations between attachment anxiety and cognitive, emotional, behavioural responses were seen across contexts; however, associations with emotional and behavioural (but not cognitive) responses were moderated by context, such that they were amplified in distress-matched online scenarios. These results suggest that social media represents a distinct context, with implications for computer-mediated communication and wellbeing links.

Whilst research suggests online social comparisons may be detrimental to wellbeing, little is known about the underlying temporal dynamics. Here we used Ecological Momentary Assessment to sample 100 undergraduate students’ experiences five times per day for 21 consecutive days, in conjunction with network analysis, to map dynamic interactions between (upward) online and offline social comparisons and multiple indicators of wellbeing. Contemporaneous, temporal, and between-subjects networks were estimated. Whilst online comparisons predicted lower self-esteem in the contemporaneous network, online comparisons predicted subsequent increases in positive, and decreases in negative, affect. In contrast, associations between offline social comparisons and poorer wellbeing were seen in all networks, and for multiple indices of wellbeing. Consistent with a two-step model of social comparisons, the findings suggest the effects of online comparisons may operate differently at different times-scales, and further, that offline comparisons may be more strongly related to poor wellbeing, with a potential causal association.

According to the Uses and Gratifications theory and Transformation Framework, social media users are drawn to different platforms according to platform affordances and motivations for use, with potential implications for wellbeing and functional outcomes. However, most research uses single- or cross-platform data. We tested the hypothesis, therefore, that the use of different social media platforms differentially predicts outcomes. Using undergraduate survey data (n = 3500+) regression analyses explored associations between time spent on eight common platforms and perceived stress and GPA scores. Platforms were also rated on design features using the Transformation Framework in order to identify potential affordance-outcome links. Our hypothesis was supported: platforms showed differential patterns of association (positive and negative) with stress and GPA, with little overlap in patterns of association with the two outcome variables. These findings suggest platforms should not be treated as a homogenous phenomenon, and implicate independent mechanisms underlying social media, wellbeing and attainment links.