Publication

Abstract : Collocated collaboration in blended settings involves the usage of technology in addition to face-to-face interactions among participants (Martinez-Maldonado et al., 2017), enabling interactions across physical and digital spaces. LA solutions often rely only on interactions captured in the digital space, offering a partial picture of the learning behavior (Pardo & Delgado Kloos, 2011). Aiming to overcome these limitations, CoTrack-has been developed to offer teachers more holistic information about student’s collaboration behavior across physical and digital spaces. CoTrack is based on Raspberry Pi and contains two components for data collection purposes: one for tracking students’ digital activity in a real-time collaborative editor (Etherpad), and another for collecting privacy-preserving audio data (it only collects the sounds’ direction of arrival, DOA). NTP (Network Time Protocol) is used for handling the time synchronization among the data collected from multiple students’ groups, and MQTT (Message Queuing Telemetry Transport) is used for transmitting audio DOA data to a server in real-time. The data collected is then analyzed to build an interaction network that shows “who talks to whom” relationship among students. It was also analyzed to obtain statistics about speaking time and writing activities. Finally, we developed an interactive dashboard for teachers to visualize the data. These features are shown for the entire duration of the collaboration activity and also allow the teacher to change the visualization for different time-frames (eg 30 sec, 60 sec, 5 min, etc). CoTrack has been used in an authentic classroom settings involving 1 teacher, 2 researchers and 9 students arranged in groups of 3. We faced the data quality issues because of the classroom noise and students’ movement. However, according to the teacher, CoTrack could be helpful for assessing collaborative activities.