Modeling with Mediation Analysis

Ran Wei,Ven‐hwei Lo
DOI: https://doi.org/10.1093/oso/9780197523728.003.0008
2021-01-01
Abstract:Building on multiple regression results, this chapter examines the direct and indirect effects of societal-level factors as exogenous variables on the studied endogenous variables to better understand the mediation effects of individual-level variables. Six structural equation modeling (SEM) models were proposed and tested. The findings clarify the role of surveillance motivation and expectancies of mobile of personal value as key mediators in affecting the behavior of mobile news consumption when societal influences (e.g., city of residence and smartphone ownership) are present in structural relationships. Similarly, level of freedom of the press as a societal factor first influences perceived utility and perceived appeal of mobile news and then indirectly influences mobile news credibility, while the effects of perceived utility and perceived appeal are derived from press freedom. Finally, information accessibility at the societal level was found to enhance learning from mobile news by directly and indirectly influencing motivations, consumption behavior, and engaged consumption.
What problem does this paper attempt to address?