Informing Pedagogical Action
Aligning Learning Analytics With Learning Design
First Published March 12, 2013 research-article
This article considers the developing field of learning analytics and argues that to move from small-scale practice to broad scale applicability, there is a need to establish a contextual framework that helps teachers interpret the information that analytics provides. The article presents learning design as a form of documentation of pedagogical intent that can provide the context for making sense of diverse sets of analytic data. We investigate one example of learning design to explore how broad categories of analytics—which we call checkpoint and process analytics—can inform the interpretation of outcomes from a learning design and facilitate pedagogical action.
via Pocket http://ift.tt/2nh0g1k
The basic premise of the article is:
Why do we need this framework?
To date, learning analytics studies have tended to focus on broad learning measures such as of student attrition (Arnold, 2010), sense of community and achievement (Fritz, 2011), and overall return on investment of implemented technologies (Norris, Baer, Leonard, Pugliese, & Lefrere, 2008). However, learning analytics also provides additional and more sophisticated measures of the student learning process that can assist teachers in designing, implementing, and revising courses (p. 1441)
Within learning design, research approaches such as focus group interviews are often used to inform redesign of courses and learning activities. The authors suggest that using analytics overcomes data inaccuracy that can be associated with focus group style research, as such approaches are reliant on self-reporting and accurate recollection of details by participants. However, they note that the interpretation of LA data against pedagogical intention is challenging, and propose a framework – “check-points and processes analytics” – for evaluating learning design.
Check-Points and Processes Analytics
In the proposed framework, two types of analytics (illustrated in the diagram above by circles and crosses in the final column) are utilised:
- checkpoint analytics“the snapshot data that indicate a student has met the prerequisites for learning by accessing the relevant resources of the learning design” (p. 1448)This type of data can be used during course delivery to ascertain whether learners have accessed the required materials and are progressing through the intended learning sequence, and prompt ‘just in time’ support (reminders, encouragement) when learners have not engaged in any required steps.
- process analytics“These data and analyses provide direct insight into learner information processing and knowledge application (Elias, 2011) within the tasks that the student completes as part of a learning design.” (p. 1448)
Again, this data could support interventions when students are involved in group work, for example, if patterns of interaction diverge from the intended patterns (unequal participation, for example, through social network visualisation).
On the one hand, this application of LA interests me because it puts LA into the work that I do as a teacher rather than at an institutional level. It feels more ‘real’ in that its focus is on pedagogy rather than the broad strokes of ‘student experience’. The institutional use of LA can sometimes seem to frame teachers as service providers and reflect the commodification of education. In contrast, this application seems like a teaching tool (with the caveat that the check-points analysis may be seen to adopt a transactional view of learning). However, I’m cautious because any plan to monitor and direct patterns of interaction is underpinned by assumptions about what effective learning looks like, and the ability to automate such monitoring and intervention through LA could enable blind adherence to a particular view of learning. Of course, even without LA we use such assumptions in our teaching: in the face to face classroom a teacher monitors group work and intervenes when students seem off task or are not communicating with each other as intended. Such interactions/engagement with tasks can (as the authors note) be more difficult to monitor in online learning, and LA could be a helpful tool for teachers online, and inform task setup and choice of technological tools used. In this sense, I would be very interested in utilising the analytics approach outlined – but I would be much less interested in it being used as an evaluative tool of my teaching, if, for example, it were based on a departmental ‘ruling’ about the types of interactions deemed to be supportive of learning, and very much less interested in using it as part of student assessment, wherein students were expected to conform to particular models of interaction in order to be ‘successful’ (see, for example, MacFarlane, 2015). As with all analytics and algorithms, the danger seems to be in the application.