We study three stochastic processes, the majority vote model, the
contact process, and the asynchronous SIR model, with phase transitions in
complex networks. In theory, we use the heterogeneous mean field theory
(HMF), which considers vertices with the same degree as equivalent, to predict
the transition points in each model and the dependence of the order parameter
on the number of vertices in the network. In each model, it was discovered that
this dependence is given by a scaling relationship like a power law. The
exponents related to the transition point deviation and the order parameter, 1/ν
and β/ν, respectively, presented values compatible with the mean field critical
exponents, presenting heterogeneous scaling corrections for the case of power
law networks. Access to finite size dependence was achieved through the use
of external fields in the differential equations of stochastic processes. The
computational simulations carried out using the Monte Carlo method agree with
the scaling relationships obtained in each model.