OpenFiction [Blog]

Achieving localization via improved nearest approximations

Posted in Uncategorized by scarsonmsm on March 14, 2006

I’m back in the thick of questions about what format is best for publishing OERs–MIT OCW uses a lot of PDF to support the widest possible access to the content, and it’s often suggested that we should switch to a format like XML to promote ease of manipulation, so that educators can more easily adjust the content. The aim of manipulation is so educators can make the material fit their needs, and this refitting of content is currently being discussed as “localization.”

Localization is a multidimensional process, with at least four facets that I’ve been able to identify:

• Currency – Updates to reflect current developments in the field
• Technical issues – Adjustments for the technical requirements of local use
• Academic issues – Adjustments for local student educational level, aptitude, prior learning and learning style, as well as incorporation into existing content or content from other sources
• Cultural issues – Adjustments for language and cultural references

The logic with the formatting questions is that if the content were in a more manipulatable format, it would be easier for educators to make these changes. It’s a reasonable enough thing on the face of it, except that a) it’s hard to decide what is the best format for doing so, and b) every bit of effort expended on reformatting materials is effort that could have been applied to publishing something else openly. Heavy expectations for formats of OERs will also discourage a great many from publishing.

I’d like to suggest that spending less time worrying about format might actually make OER more localizable, rather than less. In using OER–theoretically–you seek the nearest approximation to your need and then adjust along the four axes above. The closer a nearest approximation is to your need, the less work you have in localizing it. If the level of effort required to localize content exceeds the level of effort required to create new content, then educators will most likely roll their own. This would suggest that the more content you make available from as many sources as possible, the smaller the distance becomes to the nearest approximation for the most users.

Format conversion of any kind makes OER creation very expensive. The more expensive it is, the fewer people will do it, and the less material they’ll be able to publish. This leads to a situation where a relatively few well-funded sources of content exist that then require, on average, a large amount of localization. Those communities furthest culturally, academically, and technically from the few sources will have the most localization to do, irrespective of format, so are most likely to left to create their own.

On the other hand, if we just take what educators create, and publish as much as we can–and encourage as many others as we can to do the same–we’ll create a better network of nearest approximations. For example, I think we’re much better off with ParisTech publishing PDF engineering content in French, rather than setting the bar for OER publication at XML, not having a ParisTech OER project emerge, and then having to translate MIT content into French. Sure it would be easier to translate MIT content to French in XML, but isn’t it better (and more democratic and equitable and multicultural) to come up with open publishing norms that encourage wide participation, so that the materials are published in French to begin with?

The down side, critics may say, is that then you have more materials of suspect quality being published. Maybe, but I’m of the opinion that the great majority of educators out there are decent, hard-working, intelligent people who create very useful content for their very local and immediate needs. I really don’t believe you get oceans of junk this way. If materials are published in association with an institutional name, then you have a good enough idea of their origin and likely quality, but more importantly the academic level of the material and the audience it was intended to serve. Content from smaller institutions will have a smaller sphere of influence than–say–MIT, but they will also be better nearest approximations for the needs of important adjacent constituencies that MIT does not serve well.

Interestingly, this also allows educators themselves to decide what formats are best for their needs, in a free-market fashion. Over time, as educators themselves move to new formats, OER will keep pace with the change very nicely.

2 Responses

Subscribe to comments with RSS.

  1. […] Instead of depending on a few highly editable resources (as does the rip-mix-burn model) this relies on a wide range of resources that don’t necessarily need to be completely flexible from an editing perspective. If every school in the world is publishing their materials in a really simple format, chances are I’ll be able to find a near approximation to my needs that doesn’t require too much contextualization. Not only that, but in pointing to the resource in situ, I am allowing the learner to see the argument for which the resource was originally created, plus potentially (extending the model to include trackbacks) links to a whole range of ways in which the resource has been used. In other words, in a model that supports a multiplicity of divergent views. […]

  2. […] David uses the “scratch the itch” metaphor in discussion OER, and I think it can be applied to educational resources in general, open or not. The problem, which I think he identifies in his own way, is that with open source software, we are all largely itching in the same way (With Firefox, I believe its called “Explorer-itis”), but with educational resources, we are largely itching in our own particular way. Educational systems don’t run on the same operating system, and often don’t share the same protocols, so it’s difficult and unwise to try to converge on a single solution. I’ve suggested before that an effective educational resource is likely to be effective only for a small group of teachers and learners sharing a significant amount of context. This leads me to a conclusion similar to David’s: We need 500 more Introduction to Programming courses with some personality and some attitude. … When there are real choices out there (not “choices” between 17 clones that can’t be differentiated), then people will start finding OERs they like. This will lead to people reusing OERs. […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: