Learning analytics messages: Impact of grade, sender, comparative information and message style on student affect and academic resilience

Learning analytics enable automated feedback to students through dashboards, reports and alerts. The underlying untested assumption is that providing analytics will be sufficient to improve self-regulated learning. Working within a feedback recipience framework, we begin to test this assumption by examining the impact of learning analytics messages on student affect and academic resilience. Three hundred and twenty undergraduate students completed an online survey and were exposed to three randomly assigned learning analytics alerts (High Distinction, Pass, and Fail grades). Multivariate analyses of variance indicated significant differences between grade levels (large effects), with higher positive affect and lower resilience in response to High Distinction alerts than Pass or Fail alerts. Within each hypothetical grade level, there were no differences in student affect and academic resilience. Based upon systematic changes in feedback sender, message style or whether comparative peer achievement was included or not. These findings indicate that grade level has the largest impact on both affect and academic resilience. The failure of message and sender characteristics to impact on activities that promote self-regulated learning suggests we need to look beyond these characteristics of individual messages to identify drivers of engaging students in self-regulated learning.