Learning Analytics Dashboard Design: Workplace Learner Preferences for Reference Frames in Immersive Training in Practice

BackgroundLearning analytics dashboards are increasingly being used to communicate feedback to learners. However, little is known about learner preferences for dashboard designs and how they differ depending on the self-regulated learning (SRL) phases the dashboards are presented (i.e., forethought, performance, and self-reflection phases) and SRL skills. Insight into design preferences for dashboards with different reference frames (i.e., progress, social, internal achievement and external achievement) is important because the effectiveness of feedback can depend upon how a learner perceives it.ObjectiveThis study examines workplace learner preferences for four dashboard designs for each SRL phase and how SRL skills relate to these preferences.MethodsSeventy participants enrolled in a chemical process apprenticeship program took part in the study. Preferences were determined using a method of adaptive comparative judgement and SRL skills were measured using a questionnaire. Preferences were tested on four dashboard designs informed by social and temporal comparison theory and goal setting theory. Multinomial logistic regressions were used to examine the relationship between dashboard preferences and SRL.Results and ConclusionsResults show that the progress reference frame is more preferred before and after task performance, and the social reference frame is less preferred before and after task performance. It was found that the higher the SRL skill score the higher the probability a learner preferred the progress reference frame compared to having no preference before task performance. The results are consistent with other findings, which suggest caution when using social comparison in designing dashboards which provide feedback. What is already known about this topic? Learning analytics dashboards use visualisations to provide feedback on learning tasks to optimise learning. Learners can better understand the meaning of their feedback if it is presented alongside a point of comparison, such as a prior level of performance, the performance level of their peers or how their performance level compares to an achievement goal. Learning analytics dashboards can support learning behaviours before, during, and after learners perform a task. Learners can acquire information to improve their learning via learning analytics dashboards, such as feedback on performance, which illustrates areas of a task learners are stronger and weaker at, which in turn can help inform future training efforts.What this paper adds? Workplace learners typically prefer dashboards which offer visualisations comparing their current performance level with past levels of performance. Comparison with peers is typically the least preferred point of comparison when offered in learning analytics dashboards. No clear preference emerged between reference frames containing assigned or self-set goals in dashboards presented before and after task performance.Implications for practice Designers should take into account learner preferences when designing learning analytics dashboard visualisations. Designers should consider presenting learning analytics dashboards before, during, and after task performance. Designers should gain more insight into how learners process learning analytic dashboards and act upon it.