H817 – 20b – Week 23 – Activity 16 – Socialised learning analytics
Return to the article that you began reading in Activity 15:
Ferguson and Buckingham Shum (2012), Social learning analytics: five approaches.
Finish reading the article and make notes on the different types of socialised learning analytic and how they might be implemented.
Social learning network analytics
The Ferguson et al. (2012) paper notes that Social Learning Analytics (SLAs) may be implemented using the (now defunct) SNAPP tool (Social Networks Adapting Pedagogical Practice). This tool produces visualisations of forum contributions to help identify the levels of participation among the students.
The SocialLearn implementation is intended to provide suggestions of web pages and forum threads to an individual based on their search terms.
Social learning discourse analytics
This type of analytics is planned to be implemented using the Cohere tool on the SocialLearn platform. This tool is planned to use certain keywords in a discussion to map reasoning, evaluation, extension and challenge interactions. This data will be used to show the participant an overview of their attitudes to the subject as described by their use of these words. The tool will also show where they may be lacking in certain types of responses. In the example the participant has used less challenging language than the other three categories.
The tool could be used to identify other groups of learners considering the same subjects but approaching this in different ways.
Social learning content analytics
This type of analytic is implemented using SocialLearn’s ‘backpack’ tool. This tool allows a user to click a light bulb on any web page then identify other related resources which may be of user to the participant. It is planned that the backpack tool will later be refined to use user recommendations to improve the quality of the suggestions give.
Social learning disposition analytics
This analytic type is being implemented using a web questionnaire called ELLI (Effective Lifelong Learning Inventory). This tool generates a spider diagram which uses various metrics to display the participants level of seven different metrics as the individuals “Learning Dispositions”. This displays levels of creativity, critical curiosity and more. A summary is provided giving the learner tips on which areas of their disposition they should concentrate on in the coming months.
The questionnaire identifies areas where the learner feels they are stuck then follows through on the recommendations later to see if the learner implemented the suggested changes.
Social learning context analytics
Context analytics use factors such as location, disposition analytics and activity to provide context specific recommendations to the learner. The example given is that of a Climate Change student who is shown some recommended reading when visiting a coastal area suffering from erosion problems. This tool could provide valuable ideas for subjects to reference and look up as reading outside of their course material.
Consider the mock-ups of different learning analytics that are presented in Figures 1–5 of the paper. If you had to prioritise the development of one of these for use on the H817 module website, which would it be, and why?
Having read Ferguson et al. (2012), I would prioritise the social learning network analytics (SNAs) method. This seems to be most relevant to H817. This is because most of the interactions we are asked to provide are through the tutor forums. As I noted in activity 14 this method of analysis shows how all of the participants are interacting. However, there may need to be an element of discourse analytics introduced because the contents of the discussions are not considered using SNAs. Exploring the dialogue within the forum posts should provide the participants with important information about their contributions as shown in figure 1 in the paper.
Notes to self:
NOTES
Analysing the social media footprint of a learner would certainly require some kind of pre-permission. This type of broad data collection could be problematic from a GDPR perspective. It may be possible to implement some form of analytic which accepts social media user ID entries.
A possible implementation could be done by a data gathering program which retrieves all of the public post information for a particular user and the IDs of any users they are contacting and who contact them. This would provide an outline Social Network Analysis (SNA) of the inter-connectivity of the individual with their peers.
There may be a requirement to tag posts with a particular hash tag to allow unrelated posts to be filtered out. Perhaps language filtering could be employed but this would be very complicated.
References:
Ferguson, Rebecca and Buckingham Shum, Simon (2012). Social learning analytics: five approaches. In: 2ndInternational Conference on Learning Analytics & Knowledge, 29 Apr – 02 May 2012, Vancouver, British Columbia,Canada, pp. 23–33.