OLDaily

How Machine Learning Is Easing OER Pain Points
David Raths, Campus Technology, 2017/10/04

One of the issues I have with so many education technology projects, especially in the area of OER, is the reduction of the education problem to a search or discoverability problem. That's mostly what's happening here, and mostly how the application of machine learning to education is characterized generally. The holy grail of these applications seems to be Netflix. But we don't want to simply watch or consume learning resources, we want to do things and create things. But even thinking is treated as a search problem in this article: "If you are thinking about a topic, the machine can say, 'well based on that, have you thought about x?'" No, no, that's not how we think and learn.

[Link] [Comment]

New Quality Measurement Initiatives
Alex Usher, Higher Education Strategy Associates, 2017/10/04

Interesting article describing to recent approaches to evaluations of institutional quality: one from the Netherlands called "Measuring and Comparing Achievements of Learning Outcomes in Higher Education in Europe" (CALOHEE) which (interestingly) assesses each institution measures outcomes according to the objectives set at each institution. The other, from the United States, is the A Multi-State Collaborative to Advance Learning Outcomes Assessment (MSC) which "samples of ordinary student course work are scored according to various rubrics designed over a decade or more (see here for more on the rubrics and here for a very good Chronicle article on the project as a whole)." Alex Usher also asserts, incorrectly, that "how none of it (measurable accountability) is happening in Canada, which is demonstrably false.

[Link] [Comment]

How Communities of Inquiry Drive Teaching and Learning in the Digital Age
Terry Anderson, Contact North, 2017/10/04

Terry Anderson revisits the Community of Inquiry (CoI) model and looks at some more recent suggestions for extensions and revisions in this paper (16 page PDF) produced for Contact North in advance of the upcoming World Conference in Distance Learning in a couple of weeks (yes, I will be there). It's one of a series of insight reports being produced for the conference (yes, I contributed one, to be released soon). "Humans evolved in groups (mostly families and larger kin and tribal groups)" writes Anderson, "and these have evolved to create the social glue that facilitates learning and enhances motivation in the COI model. The continuing popularity of the model, through different technologies, shows that group based learning is still highly valued and the most common way in which at least young people engage in both formal and informal learning."

[Link] [Comment]

Is it time for a Photoshop button on Instagram?
Isaac Fanin, BBC News, 2017/10/04

There is discussion this week about the role of Photoshop after a new French law stipulates that "it will be mandatory to use the label ‘retouched photo’ alongside any photo used for commercial purposes when the body of a model has been modified by an image-editing software to either slim or flesh out her figure." In this post British actress Mel Wells argues that the thinking should be extended. ""The aspirational body types are just not realistic for 95% of the population and because of that it's really damaging people's self-esteem." But almost all photographs are altered, without an editing history it's difficult to tell whether a body has been reshaped or whether a blue hue created by the camera has been corrected. But it's worth keeping this in mind: advertising is the original fake news.

[Link] [Comment]