One of the things we’ve talked about in SoLAR recently is the potential of established platforms for learning analytics data. Two of the most used platforms are Mediawiki (which Wikipedia is on) and WordPress (which this blog, along with a huge number of others is on). Both of these allow extensions or plugins (respectively) to add functionality, or get at data which is otherwise difficult to probe (manually). They both also have APIs from which a range of potentially useful data could be gathered. Given the wide use of these platforms by learning institutions already, it seems to me that some guidance on how you might use data from them to get meaningful insight would be quite useful – does anyone know if anything like that’s been written? From a quick search of the Mediawiki extensions list, I can see a few things I can imagine using (as below) – although of course, it always depends on the task being set, the sort of data you want to get out (and how you use it) and so on. Some Mediawiki extensions (many experimental) I can imagine being useful: * WikiTrust http://wikitrust.soe.ucsc.edu/ – useful to explore ‘reputation’ models of trust amongst students in shorter term assignments, and in longer term work to provide simple metrics which might be used for reputation to confer particular benefits on some users – for example, allowing more established and high reputation users to give badges to newer users (if you have more ideas on this, Doug Belshaw of the Mozilla Foundation was recently seeking advice here: http://dougbelshaw.com/blog/2013/01/27/some-brief-thoughts-about-online-peer-assessment/ ) * SearchLog http://www.mediawiki.org/wiki/Extension:SearchLog which logs use of the search bar within a Wiki, could be used to explore concept navigation by students, provide data for linking your own content together, create concept maps from search trails, provide suggestions for areas of interest (although this would need to be a fairly large wiki, preferably with broken down pages to work effectively) * Collaboration Diagram https://www.mediawiki.org/wiki/Extension:Collaboration_Diagram – gives a visualisation graph users’ contributions to: a given article; several articles; all pages in a category; with an edge weighting such that the user who contributes the most has the thickest line connecting them to the article node. Might be interesting alongside WikiTrust above, interesting way to see user contributions within a group and the ways that groups interact. * CountEdits http://www.mediawiki.org/wiki/Extension:CountEdits count the edits made by any particular user…yawn * Contributions http://www.mediawiki.org/wiki/Extension:Contributions not dissimilar to WikiTrust above (but with user ratings included), gathers data on “number of pages, number of edits, number of words, number of views, edit longevity and user ratings.” I can also imagine analysis on talk pages (e.g. Pifarré and Kleine Staarman, 2011), automated analysis of text for: complexity; internal links; citation; structure, etc. (see also my blog on [using wikipedia for exploring epistemic beliefs]1). So – what do people think, what other extensions would be useful? Should we be making more extensive use of tools such as this a) when trying to gather LA, and b) in our general assessments? Ref:
Pifarré, Manoli, and Judith Kleine Staarman. “Wiki-supported Collaborative Learning in Primary Education: How a Dialogic Space Is Created for Thinking Together.” International Journal of Computer-Supported Collaborative Learning 6, no. 2 (April 15, 2011): 187–205. doi:10.1007/s11412-011-9116-x.
Footnotes
-
http://sjgknight.com/finding-knowledge/2013/01/wikipedia-feedback-ratings-as-an-epistemic-tool/ “Wikipedia Feedback Ratings as an Epistemic Tool” ↩