Based on VKontakte data, we study the influence of various factors on the dynamics of opinions and actions both at the macro level (“public opinion”) and at the micro level (opinions and actions of individual agents). This paper concludes the multi-part study. Identification results are presented for binary models (threshold models and models with latent variables) that describe the dynamics of agents’ opinions and actions in a social network. These models are used to estimate the influence of various factors on agents’ opinions and actions (public opinion, the agent’s individual opinions and actions, the opinions and actions of the social environment, and the mechanisms of the agent’s trust in information sources and information content). Finally, linear models are compared with threshold models and qualitative findings of the multi-part study are drawn.