Computational and causal examinations of wellbeing in situated contexts by leveraging social media and multimodal data

Abstract

Assessing wellbeing can be complemented with social and ubiquitous technologies. This dissertation uses social media in concert with multimodal sensing focusing on situated communities. Before incorporating such assessments in practice, we need to account for confounds impacting behavior change. One such confound is 'observer effect', that individuals may self-alter their otherwise normal behavior because of the awareness of being 'monitored'. My proposed work studies this problem on social media behavior. On a multisensor study of 750 participants, I intend to conduct a causal study of modeling behavior change during study participation. This work will provide valuable insights and guide recommendations for correcting biases due to observer effect. This dissertation bears implications for social computing systems and stakeholders to support wellbeing and crisis intervention efforts in situated communities.