Opinion formation under mass media influence on the Barabasi-Albert network

Abstract

We study numerically the dynamics of opinion formation under the influence of mass media using the $q$-voter model on a Barabasi-Albert network. We investigate the scenario where a voter adopts the mass media's opinion with a probability $p$ when there is no unanimity among a group of $q$ agents. Through numerical simulation, we identify a critical probability threshold, $p_t$, at which the system consistently reaches complete consensus. This threshold probability $p_t$ decreases as the group size $q$ increases, following a power-law relation $p_t \propto q^{\gamma}$ with $\gamma \approx -1.187$. Additionally, we analyze the system's relaxation time, the time required to reach a complete consensus state. This relaxation time increases with the population size $N$, following a power-law $\tau \propto N^{\nu}$, where $\nu \approx 1.093$. Conversely, an increase in the probability $p$ results in a decrease in relaxation time following a power-law relationship $\tau \propto p^{\delta}$, with $\delta \approx -0.596$. The value of the exponent \( \nu \) is similar to the exponents obtained in the voter and $q$-voter models across various network topologies.