In the process of disease transmission, the decision of whether or not to vaccinate is not only influenced by the environment, but also by household factors. Here, we extend the susceptible–infected–susceptible–vaccinated (SIVS) model by introducing two kinds of information related to decisions and study disease transmission on household networks, where vaccination not only depends on the environment, that is, the number of infected neighbors, but also relies on the situation of household members. If a household member decides to take the vaccine, his susceptible household members will also be vaccinated unconditionally. Through extensive numerical simulations and theoretic exploration, results reveal that such a household-based collective vaccination decision cannot prevent the epidemic compared with the personal-based vaccination decision. The household-based collective decision can only protect the vaccinated families, but it cannot protect the unvaccinated families. Unlike family-based collective decisions, personal-based vaccination decisions can successfully stop the spread of disease by separating susceptible individuals from infected ones. Our study thus enhances the understanding of the evolution of epidemic about collective vaccination and personal vaccination decisions. • In this paper, we combine the scale-free network and fully connected network as the household network. • We set two kinds of household networks according to the real situation. • We consider the dynamic of vaccination coming from different information sources. • A family-based collective vaccination decision can not prevent the epidemic compared with a personal-based vaccination decision. [ABSTRACT FROM AUTHOR]