PATTERNS OF INTERACTION OF A COLLABORATIVE PROCESS FOR IMPROVING THE INITIAL DESIGN OF A LEARNING ANALYTIC

The use of activity logs that users of digital learning platforms generate by participating in these platforms has been analyzed from different lines of research; these lines try to use the enormous amount of data to improve the experiences learning of the users. In detail, the research and design field of Learning Analytics has focused on measuring, collecting, analyzing and presenting data about activity logs of the users for the purpose of understanding and improving of learning collaborative processes and the environments in which are performed. However, different works point out that Learning Analytics must combine multiple levels of analysis to provide a deeper insight into the collaborative learning process and improving the activity of users of digital learning platforms. On this basis, a research was carried out to analyze if the multilevel information provided to the participants in a digital forum has an impact on their collaborative processes; for this purpose, 24 undergraduate students participated in a series of online forums. The results indicate that collaborative processes of the participants that had multilevel information available improved much more than processes of the participants that just had information of a single level; however, this difference is more noticeable in the first activity forums, since it gradually decreases over the forums. In addition, the multilevel analysis provided an in-depth view of the collaborative processes of the participants, since different patterns of interaction were identified through which the participants contribute to these processes. It is concluded that any Learning Analytic must pay special attention to the characteristics of the task around which the collaborative process is articulated.