Article Preview
TopIntroduction
With the blooming development of social media, the use of social media in the field of education has recently begun to receive scholarly attention (Ranieri, Manca, & Fini, 2012). Students can share knowledge and have discussions with peers to gain new knowledge through a social networking mechanism. However, how learning is influenced in a social learning environment still needs to be investigated in detail. Much research discusses how peer interactions affect social learning (Cho, Gay, Davidson, & Ingraffea, 2007; Moreno, Gonzalez, Castilla, Gonzalez, & Sigut, 2007; Pear & Crone-Todd, 2002; Puntambekar, 2006). It was suggested that the interaction situation affects the way the students distribute knowledge, for instance, how students give and receive their knowledge (Pear & Crone-Todd, 2002), which of the members involve in the interaction (Moreno et al., 2007), how interactions happen for foster learning (Puntambekar, 2006), and whether different interaction styles affect learning (Cho et al., 2007). To understand more about the effects of peer interaction on learning, increasing research focused on trying to explore behavioral patterns of peer interaction (Liu & Tsai, 2008; Hou & Wu, 2011). Most of them analyzed peer interaction types based on statistical methods (Hou & Wu, 2011; Moskaliuk, Kimmerle, & Cress, 2012) or content analysis (Puntambekar, 2006). However, by only employing those methodologies, the ability of mining students’ static (topologies) or dynamic (distribution) interaction characteristics will be very limited. Therefore, social network analysis (SNA) grows into an emerging computing paradigm, which can mine structural and topological features from the social network by leveraging more effective algorithms (Firdausiah Mansur & Yusof, 2013; Huang, Li, & Chen, 2005).
Instead of applying traditional methodologies, Cho et al. (2007) explored the relationships between peer interaction styles and learning performance by utilizing SNA indexes. In addition, Chen and Chang (2012) applied SNA to identifying students’ interaction types in a problem-based learning (PBL) environment by deriving peer interaction indexes from their interaction situation, based on which the recommendations of learning partners for supporting collaborative PBL were provided. The developed peer interaction indexes can mine students’ behaviors more accurately and objectively. However, in addition to the partner recommendation and the PBL environment, SNA may have the potential to be applied to address more general issues in social learning. Increasingly, recent research (Firdausiah Mansur & Yusof, 2013; Lu & Churchill, 2014) has started to apply SNA to explore the effects of peer interactions on learning. However, such research about the students’ peer interactions in a general social learning environment is still lacking. Therefore, in this research, the effects of peer interactions on social learning in terms of learning achievement were investigated by employing SNA techniques. We developed SNA algorithms to analyze students’ peer interactions. Accordingly, in order to examine this issue in a more general social learning environment, a social learning platform with general social functions was carried out for examining the students’ peer interactions in social learning.