Polarized media drives opinion drift in new Deffuant-Weisbuch model study

Researchers Oliver Zheng and Mason A. Porter have developed an extension to the classic Deffuant-Weisbuch (DW) bounded-confidence model (BCM) of opinion dynamics by introducing influence from two media sources—one with a positive value and one with a negative value. In the original DW BCM, agents on a network adjust their continuous-valued opinions through pairwise interactions, but only if their opinions are within a certain confidence bound of each other. The new model captures the effects of a polarized media landscape where agents are also influenced by external media sources. Both numerical simulations and analytical derivations reveal a striking drift behavior: a large cluster of opinions systematically moves toward one of the media agents over time, even when both media are equally influential.

The study analyzes how the drift trajectory and speed depend on model parameters such as the confidence bound, media strength, and the number of agents. They identify conditions under which drift is promoted or suppressed—for instance, higher confidence bounds and stronger media influence accelerate drift, while certain network structures can slow it. The results provide quantitative insight into how competing media sources can collectively push public opinion in a particular direction, offering a rigorous framework for understanding polarization in social systems. This work has implications for designing interventions to mitigate media-driven opinion manipulation.