Lately, network sampling proved as a promising tool for simplifying large real-world networks and thus providing for their faster and more efficient analysis. Still, understanding the changes of network structure and properties under different sampling methods remains incomplete. In this paper, we analyze the presence of characteristic group of nodes (i.e., communities, modules and mixtures of the two) in social and information networks. Moreover, we observe the changes of node group structure under two sampling methods, random node selection based on degree and breadth-first sampling. We show that the sampled information networks contain larger number of mixtures than original networks, while the structure of sampled social networks exhibits stronger characterization by communities. The results also reveal there exist no significant differences in the behavior of both sampling methods. Accordingly, the selection of sampling method impact on the changes of node group structure to a much smaller extent that the type and the structure of analyzed network.