Using stochastic actor-oriented models to explain collaboration intentionality as a prerequisite for peer feedback and learning in networks

Peer feedback and collaboration intentionality (CI) are key prerequisites to advance in higher education. For learning, it is crucial that peers do not merely interact, but that students are willing to function as scaffolds by sharing their knowledge from different perspectives and asking each other for academic support. Peer feedback can only take place within a collaborative learning approach and when students are willing to initiate feedback relationships with their peers. Therefore, we analyze peer feedback networks (in terms of academic help and advice-seeking) and CI as an individual characteristic using an advanced statistical tool, namely stochastic actor-oriented models (SAOMs). In SAOMs, we control for selection and influence mechanisms. Selection comprises instances when feedback relations can be initiated based on CI, while influence builds upon existing feedback relations in affecting CI. One important selection mechanism is homophily, which means that individuals prefer to initiate a connection with someone else based on similarity in characteristics, attitudes, or behavior. In this chapter, we introduce this statistical technique within the higher education context and the added value for feedback research in education. We illustrate the SAOM methodology using two-wave peer feedback networks and CI data while controlling for gender and the Five-Factor Model personality traits. In this empirical example, we address the research question: To what extent does homophily of CI plays a role in selecting peers when seeking feedback and to what extent do feedback relationships influence CI? The SAOM shows an homophily effect, which implies that students preferentially seek feedback from others who are similar in CI. We also find an influence effect in which students who seek feedback from one another become more similar in terms of CI over time. Similarity in CI is driven by selection and influence mechanisms in peer feedback networks.