Text: aaa | Text only

Wikipedia in Education, and tracking how 'knowledge' is used...

On: 16th September 2012

EduWiki Conference & the power of 'paradata'

Last week I went to the EduWiki conference  in Leicester.
Twitter EduWiki hastag wordcloud

Most of the speakers talked about use of Wikipedia in the context of asking students (of whatever age) to edit articles.  Sometimes this was subject based content, others it involved involved translating from one wiki to another language as a language teaching method.  This educational use was also considered with regard to accreditation/assessment. There are a number of projects out there doing interesting things, to give a taste

  • Chris Trace talked about Wikivet - based at the Royal Vetinary College– a wiki for vetinary medicine, only editable by qualified practitioners, used around the world in multiple languages.
  • A few people mentioned the GLAM (Galleries, Libraries, Archives and Museums) program working to maximise impact of collections online
  • Cornelia Trefflich from WikiMedia Deutschland talked about their Silver Knowledge (Silberwissen) program (see google translation here).  This program is focussed on enabling older users (particularly those who are underrepresented, but still working – in this case targeted at the over 50s) to edit Wikipedia.  This program has encouraged such users to bring their wealth of knowledge, both local and subject specific, to Wikipedia. 
  • Not talked about at the conference, but another interesting project will hopefully be actioned this year to improve the Sex Education resources on Wikipedia, in recognition that while it is widely used for information, these resources are not always of the highest quality. 

As well as these projects, some other interesting ideas were talked about in terms of education, and evaluation. 

1) Doug Belshaw (of the Mozilla Foundation's Open Badge project) was on a discussion panel talking about Badges and accreditation of learning (see Bob's and my Nominet Trust blogs)

2) Amber Thomas (of JISC) talked about paradata -  

As a general definition, paradata are usage data about learning resources that include not just quantitative metrics (e.g., how many times a piece of content was accessed), but also pedagogic context, as inferred through the actions of educators and learners http://en.wikipedia.org/wiki/Paradata_%28learning_resource_analytics%29.

Why is paradata important?

Quantiative metrics are often the focus of evaluation -  patients seen, website hit counters, grades or scores achieved, and so on. 

Qualitative analysis - interviews story data, increasingly multimedia (images, videos, blogs) are important additions for understanding the real world context of whatever we're evaluating

Paradata could provide a system to allow us to explore the ways information is being used, in large scale quantitative ways which goes way beyond the most simple quantitative metrics.

For example, on my current project (ORBIT), we’re using a template to keep track of the ways segments of larger documents are being used in various other places across the wiki; which sections, and what type of document they are used in tells us something interesting.  We’re hoping to see which documents are most widely used in combinations, which bits of them, and whether they’re present on ‘article’ style pages, activities, introductions, etc.

Using Paradata

But this isn’t really enough – paradata should extend beyond an individual site, particularly when the materials are Creative Commons licensed and can be widely reused.

Imagine, you’ve created some data, or document and followed the 5 star open data guides:

* make your stuff available on the Web (whatever format) under an open license1

**make it available as structured data (e.g., Excel instead of image scan of a table)2

***use non-proprietary formats (e.g., CSV instead of Excel)3

****use URIs to identify things, so that people can point at your stuff4

*****link your data to other data to provide context5

http://5stardata.info/

Imagine the power of being able to see – in an open, standardised way – how your data is being used across the web. 

  • What sort of claims does your data appear alongside?
  • What sort of materials does your data appear in?  What is the tone of these materials?  What is the purpose (instructional, activity based, ‘presentational’, argumentative, synopsis, etc.)
  • Which parts of your data/materials are being most widely used, and how?
  • Are there other organisations/people whose materials often appear alongside yours, and could you partner with them?

Conclusions

Paradata provides an interesting addition to the analytic toolkit which I'd love to explore further, but I do have a few notes of caution:

  1. For it to be widely understood, and have the most impact, a standard, and reliatlvely easily implementable one at that, needs to be in common use
  2. Deployers still need to be aware that (obviously) the content they're tracking is still hugely important - this isn't a substitute for well founded, researched, and consulted on content prior to deployment.
  3. It also isn't a substitute for either other methods of analysis (including quantitative metrics, or qualitative stories gathered from end users), or the 'theory' which facilitates our sensemaking of such data, our trying to understand the data in real world contexts.

I'd love to hear about it if anyone's explored paradata (or similar) in their context.  I can see its use in traditional, and informal learning contexts, but I can also imagine it being hugely powerful in other fields - including for example, the MediaWiki project to improve their sex ed resources, and similar endeavours.  As ever, get in touch on twitter or below the line.

(For the interested reader, there's an EduWiki tweet archive here and a Storify of day 1, and day 2).