Learning Analytics

In "In-person Coaching Versus Technology: Proactive, Constant Contact Matters," Oreopoulos & Petronijevic (2016) compare the relative effects of self-paced online training, generic email/SMS messaging campaigns, and personalized one-on-one coaching on student GPAs. Their results indicate that the moderating effects of trust and rapport between students and their advisors may not translate or scale well to lower-cost, technology-mediated approaches.

How might learning analytics shape our experience of education itself? Swan suggests four ways: epistemologically, ontologically, systemically, and politically. Epistemologically, learning analytics may influence what kinds of knowledge, learning, and teaching are prioritized based on that which can be quantified, tested or otherwise measured. Ontologically, learning analytics will shape the 'reality' of education and learning by privileging practices and policies from which quantifiable outcomes can be derived - with implications for the role of human reason in the learning process. Learning analytics may shape education systematically in that quantifiable assessments or metrics can be implemented at every level and department of institutions. Politically, learning analytics shape education by both requiring and legitimizing certain forms of governance and resource allocation.

Referenced in Swan (2016), Dringus lays out the functions learning analytics must support in order to minimize harm to students. These include data accuracy and validity, transparency, sound algorithmic processing and analysis, thoughtful and effective assessment, and use to inform processes and practice.

"The New Version of Administrative Creep" (2017) continues a thread from Swan (2016) - the idea that learning analytics will shape education ontologically; in other words, redefine the scope of education and educational practices. Carlson & Filner don't discuss learning analytics outright, but rather the phenomenon of faculty adopting an administrative mindset when considering issues in the academy. Are there parallels?

June takes us to UNC Greensboro (also a Starish outfit) in "How a University Fights to Keep Students' Demographics from Becoming Their Destiny." UNCG is taking steps to actively manage the disparate impact of learning analytics on students with the greatest learning needs.

Nelson provides a case study of nurses' experiences with electronic health records. Of interest to our discussion are the promises made about EHRs' benefits for patient care, ease of use, and security, in comparison to nurses' real experiences. How does this compare to the current state of learning analytics?