Learning Analytics

Core Questions

What is learning analytics?

What is the difference between educational data mining on the one hand, and learning analytics on the other hand? Is the perspective in Siemens and Baker more compelling, or the perspective in Ferguson? What are both perspectives missing, if anything? (Let's look at the claims one by one)

In past classes, I have claimed that learning analytics is associated with holism and with McKeon's ontological perspective. Is this claim reflected in these articles?

How do the Ferguson and Siemens articles differ in writing style from the three EDM articles? Is this difference reflective of anything deeper?

What is the difference between learning analytics and academic analytics? Are they properly the same field or different fields?

Let's discuss the Learning Analytics Model in Figure 2 of Siemens (2013).

Secondary Questions

Siemens discusses a tension between innovation (generating something new) and analytics (evaluating what exists in data). Is this tension real? Can we generate genuine innovation from data?

Siemens (2013) argues that quality data is important, as is interoperability, whereas Baker has talked about the importance of being able to get something valuable from "roadkill" data. What are the relative merits of these two perspectives?

Siemens refers to learning analytics as a field... but unlike the EDM articles and the Siemens article, Ferguson refers to learning analytics as an area of technology-enhanced learning with roots in a variety of fields. Aside from the obvious (look at the journal name), what are the roots and implications of conceptualizing learning analytics this way?

The Ferguson article cites social network analysis as a key step towards socially driven pedagogies entering learning analytics/EDM. Why is this a particularly powerful method? What are its relative benefits in comparison to discourse analytics methods for studying collaboration?

The Ferguson article explicitly discusses the political factors driving the growth of learning analytics. In the previous class on EDM we discussed political factors in the USA that may hamper the uptake of this field. What are the factors driving, slowing, and shaping LA/EDM in the USA and elsewhere? Do these factors differ by country?

The list of tools mentioned by Siemens barely overlaps at all with the tools used in EDM, either the EDM-specific tools or the general tools (RapidMiner, Weka, and KEEL). What are the implications?

How much of a problem are the internal limitations that Greller and Draschler discuss? How can this be most effectively surmounted?