Many people feel their emotional wellbeing is affected by their social media use, in positive and/or negative ways. However, scientific research has made little progress in uncovering the mechanisms by which social media could affect mental health. We now know that social media use can be modelled as a reward-learning process, whereby people engage with social media partly in order to gain ‘rewards’ such as ‘Likes’ and followers. However, we do not know which individual differences cause people to pursue and react to these rewards differently, nor whether different responses to these ‘rewards’ might determine the effects of social media on wellbeing.

We propose to investigate, for the first time, reward-learning behaviours as a potential mechanism linking social media and mental health. We intend to use the unique Twitter-data linkage ALPSAC has established for its participants in recent years. We will link the multidimensional questionnaire data regarding interindividual differences and developmental trajectories in the ALSPAC cohort, to data from these individuals’ Twitter profiles. We will then use computational models of reward-learning to model the Twitter data. This will allow us to associate behavioural and emotional trajectories with reward-learning behaviours on Twitter.

We will attempt to answer the following questions:
- Are behavioural and emotional characteristics, and other individual differences, associated with reward-learning behaviours on Twitter?
- Do reward-learning behaviours, such as the extent to which people alter their posting in response to the numbers of ‘Likes’ they receive, predict mental health trajectories over subsequent years?