Please use this identifier to cite or link to this item:
|Title:||An Alternate Statistical Lens to Look at Collaboration Data: Extreme Value Theory|
|Publisher:||International Society of the Learning Sciences (ISLS)|
|Citation:||Sharma, K. & Olsen, J. (2019). An Alternate Statistical Lens to Look at Collaboration Data: Extreme Value Theory. In Lund, K., Niccolai, G. P., Lavoué, E., Hmelo-Silver, C., Gweon, G., & Baker, M. (Eds.), A Wide Lens: Combining Embodied, Enactive, Extended, and Embedded Learning in Collaborative Settings, 13th International Conference on Computer Supported Collaborative Learning (CSCL) 2019, Volume 1 (pp. 400-407). Lyon, France: International Society of the Learning Sciences.|
|Abstract:||To provide beneficial feedback to students during their collaboration, it is important to identify behaviors that are indicative of good collaboration. However, in a collaborative learning session, students engage in a range of behaviors and it can be difficult to indicate which of those behaviors correlate with higher outcomes. In this paper, we propose using Extreme Value Theory (EVT), a method that considers the data points in the tail (upper or lower) of the distribution, to analyse the relationship between collaborative process variables and outcome measures through insights derived from high impact, low-frequency events. Specifically, in this paper, we analyse the relationship between dual gaze patterns and outcome measures across two different datasets. In both datasets we found that students with lower outcomes had lower focus during the collaborative session. This paper provides a contribution by both introducing EVT as a viable method for analysing CSCL data as well as demonstrating the effectiveness of eye- tracking as a collaborative indicator to use to adapt to in real-time.|
|Appears in Collections:||CSCL 2019|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.