Code development has been performing collaboratively for a long time. Platforms, such as GitHub, contribute to this process with various mechanisms. Pull Request is a mechanism that allows developers to submit their contributions to a project. Then, these changes can be discussed, analyzed, and reviewed before being integrated into the repository. One of the goals of this process is to avoid a phenomenon called design decay. It occurs when poor design structures are introduced in a project. As a result, the project may become difficult to maintain and evolve. Existing techniques use source code symptoms (e.g., code smells) to identify the manifestation of design decay. Nevertheless, such symptoms can only be used to identify design decay that is already present in the project. Thus, in this dissertation, we investigated the exploration of three social aspects to predict the manifestation of design decay on open-source projects as follows. Communication Dynamics represents information about contributor s roles and temporal aspects of their discussions. Discussion Content is the information being exchanged among participants of a contribution. Finally, Organizational Dynamics represents characteristics of the team organization. The manifestation of these social aspects along software development can induce behaviors that possibly affect the design quality. However, no previous study has investigated the influence of such social aspects on the manifestation of design decay. Thus, we aim to shed light on how these three aspects influence the design decay. To achieve this goal, we introduced a suite of metrics for characterizing social aspects in pull-based software development. Then, we analyzed seven open-source projects, mining both their commits and pull requests. Our results reveal that: (i) many social metrics, e.g., Discussion Length, can be used to discriminate between pull requests that impact on the manifestation of design decay from the ones that do not impact; (ii) various factors of communication dynamics, such as Number of Users, are related to design decay. Nevertheless, temporal factors of communication dynamics outperform the participant roles as indicators of design decay; and (iii) aspects related to organizational dynamics, such as the number of newcomers, are surprisingly not associated with design decay manifestation.