In this study, we investigate the role of the mesoscopic structural properties of a scale-free social network on the contagion spreading. We focus on both the exponent of power-law community size distribution function (β) and the mixing parameter (μ). Findings show that increasing β reduces the rate of epidemic spreading. On the other hand, increasing μ increases the rate of epidemic spreading. Two innovating parameters, Temperature and cos θ, are introduced here to analyze these effects.