Scholars have examined how students feel, behave, and think when engaged in asynchronous online discussion (AOD), which has become a prevalent means of collaborative learning in higher education. However, few have considered the impact of time on these learning-related dimensions, partly due to reliance on self-reported instruments that are not sufficient to measure the temporal, changeable, and iterative relationships among learning-related dimensions during online conversation. The purpose of the current study was to use learning analytics to investigate longitudinally the relationships among learning-related emotions, learner participation level, and cognitive effort of students. Analysis of system-stored student data from 56 students enrolled in a wiki-based, blended undergraduate writing course revealed that positive or negative expressed emotions led to higher participation in ongoing discussions and that participation level predicted the rate of change in textual revision. These findings not only confirm theories of learner engagement but also demonstrate the viability of trace data and their potential in educational research and practice.

No abstract