This entry features a cognitive profile of data available in the Facebook Open Graph API. In this first part of a series of “Thinkstitching” entries, I’m using the Dunn and Dunn Learning Styles Taxonomy as a means of measuring the range of cognition that may be measured with these data.
Dunn and Dunn are, for me at least, the easiest place to begin cognitive mapping of an app or an API because its taxonomy can be used to describe inputs or evidence of behaviors when analyzing interactions by data types rather than by learners or learning outcomes. For details on the model, have a look at a PDF from the UK firm, seechange consulting: http://www.seechangeconsulting.com.au/images/Page_file_Library_Documents/Dunn%20and%20Dunn%20Model.pdf
The value of this type of mapping depends on the audience. Primarily, the work is meant to enable better analytics on vast data sets by providing some clues on data types and combinations of them that can demonstrate learning or growth. Even if an analytics initiative is not aimed at education or Human Capital Management, the structures in instructional activities are useful for mining interaction models and growth/development within them. Also note that inputs to these initiatives assume that the data are meant to represent archetypes or aggregations rather than individuals. Data mining to inform profiles of individuals is certainly possible, but it is out of scope for this particular blog entry and related effort.
Product development teams and educational program administrators might also see value. Product folks might appreciate these maps as reviews on the cognitive/behavioral richness of the interactions they design. Administrators, particularly those feeling the pressure to change assessments to something that more authentically captures and measures higher order thinking skills, can see how various apps or technologies can support real-world, “multi-app” assessment activities, both in terms of delivery to learners and capture via APIs or direct access to data stores. Big picture, we don’t need to build vast new systems to support fully integrated authentic assessments. The more authentic approach would be to stitch together technologies that exist and that people already use.
Anyway, the cognitive profile of the Facebook API is below. It functions as a scorecard at this high-level, summary view. The richness of Facebook data in terms of inputs to and circumstances of cognition is the value being measured here. In the next couple of entries, I’ll provide another layer of detail by identifying specific data types from Facebook to support this entry’s summary view, and I’ll drill down further in terms of analyzing the value of this work. Later on, I’ll add analysis of more apps or APIs.