Behavioral factors may be utilized to assess effectiveness of individual learning processes and contexts in which it occurs (Muñoz-Merino et al., 2015). Most research explored students’ behavioral patterns by examining a single aspect of their interactions, such as students’ time spent on assignments (Champaign et al., 2014), students’ navigations (Guo & Reinecke, 2014), in-video dropouts (Kim et al., 2014a) and forum posting behaviors (Wen & Rosé, 2014). As for the study on multiple aspects of interactions, Wen and Rosé (2014) proposed a novel topical model to identify students’ behavior patterns using six indices (video lecture, assignment, quiz, survey, forum participation, browse the course material). Nevertheless, most of these studies did not include high-level interactive behaviors for deeply understanding students’ learning process. To sum up, these studies sought to explain the potential reasons of these behavioral diversities or establish a mapping relationship between groups with students’ achievement.
On the other hand, some research tended to explore sequential behavior patterns to find consecutive changes of behaviors within discussions and collaborative problem-solving in online learning systems (Hou & Wu, 2011; Liu et al., 2017). As for informal learning contexts, the main stream was to explore the sequential behavior patterns in knowledge construction. For instance, Shukor et al. (2014) revealed that, the high-performing students tended to present a statistically significant sequence in negotiating shared information after entering forums. With such information, the educators could decide which action strategies were critical for better assisting students to construct knowledge.
As indicated in Cerezo et al. (2016), however, most studies were still conducted in laboratories or researcher-controlled settings over a short period of time, in which behavioral observation might be invasive for students and explored results could not be generalized to the real-life formal learning environments. To address these limitations, this current study was carried out in a SPOC (being a supplement to face-to-face classroom teaching), rather than a controllable laboratory setting.