To recap, my definition of Learning analytics is

“The timeous capture and analysis of institutional data sets, to inform immediate and long term decision making.”

Having reviewed the Ferguson et al. (2019) paper, I have identified the following innovations which I believe could be supported by learning analytics. I do believe my original definition is still valid in the context of the following examples.

Innovation 1 – Action Learning

Action Learning could perhaps be supported by analytics. This learning technique could form the basis of a data analysis model.

Action Learning is attempting to recommend actions in response to a problem using the collective skills and knowledge of a selected group of individuals.

Since learners work in groups, analytics could be used to determine the best mix of background skills to combine into a team to maximise the potential of the group.

A bank of questions could be stored from past action learning groups. The success rate of solution finding based on the characteristics of the team members (and question type: STEM, creative, psychology etc.) could form the basis of a predictive algorithm.

The data will show a mixture of success and failure in solving various problems by taking certain actions. These actions may form patterns of responses which lead to success or failure. This data could be used to assist with some ‘preset actions to try’ which have proven previous success.

Learners and teachers may need support in navigating any system which has been built using previous problems. The list of possible subjects is so vast that the exact problem presented may not be stored.

Innovation 2 Learning through wonder.

The example I have some experience with would be a Blue Box event as described by [Ferguson] suggesting pedagogies for exploring a zoo. The College E-Assessment Group (CeAG) (2019) Winter Fayre programme details a “Google Forms on the go workshop”.

This workshop provided instruction on creating real-time assessment for exploration and learning away from the classroom.

The data gathered could be children’s experiences interacting with various habitats in the zoo. There may also be (as was the case in the CeAG event), assessment questions to be answered while following one of the predetermined zoo tours. Mobile assessment could be improved using quiz metrics and feedback from the participants.

This could be used to help design new experiences or provide important evidence that key curriculum indicators are being met.

The assessment designers may require support from subject experts in creating the most effective learning experiences.

Innovation 3 – Making Thinking Visible

Making thinking visible aims to allow students to use a variety of tools to express their responses to questions.

I think analytics would be relevant in the creation of new lessons based on information gathered from student thinking rather than as Ferguson identifies “Assumptions about student understanding”.

The data available will be dependent on the subject and recording medium but these could be group work outputs, blog posts or a portfolio of work produced by a student.

Teachers will require assistance in using the body of student work to design new course materials which embrace and extend the ideas found.

Although this analysis is likely to overlap with “Networked Learning” techniques it would have a greater impact on the teacher. This would be construction of lessons rather than discovering how students gather knowledge through interaction in online spaces.

Where it would be difficult to implement analytics

I did not consider Virtual Studios as a potential candidate for Learning Analytics. As an example of Networked learning, I investigated the use of learning analytics and discovered that this work is already being done. Ferguson (2012) explains how Social Learning Analytics (SLAs) are being implemented in the OU (2012) SocialLearn system.

I also think it would be difficult to implement analytics for drone teaching. The success or failure criteria to analyse through this teaching technique is not obvious. It is difficult to visualise what data could be gathered from such interventions to provide potential analytical models.

---

References:

College e-Assessment Group (2019) Winter Fayre – Live Programme, Available at: https://docs.google.com/document/d/17AVh3JHwM2d1tqB0AYiIVyiR3-uApsfdQi1_g2P4Y38 (Accessed: 6th July 2020).

Ferguson, R. and Shum, S.B., 2012, April. Social learning analytics: five approaches. In Proceedings of the 2nd international conference on learning analytics and knowledge (pp. 23-33).

OU (2012) SocialLearn, Available at: http://www.open.ac.uk/blogs/SocialLearnResearch/ (Accessed: 6th July 2020).