I reckon using softwares to generate the Analytics shouldn’t be difficult, as nowadays there are a lot of turnkeys, like the SPSS, or the Statistics Package, PROMODAL, ithink etc. The challenge is: (1) how would we interpret the statistics? (2) what does the statistics or analytics tell me (as an educator or a learner)? and (3) what actions or interactions are appropriate for certain patterns of connections?
I also think that learning analytics are like thinking that the use of SQC (Statistical Quality Control) – the 7 QC tools or even the new QC 7 tools – Quality Function Deployment, Affinity Diagrams, Relationship Diagrams, etc. might have given us all the solutions. Surprisingly, they won’t, mainly because the complexity behind all these learning won’t be that simple. Also, would lots of assumptions behind those aftermaths statistics be good predictors of the future? New agents, new interactions, and new environments would yield new results, and so only novel solutions – i.e. based on probe, sense and response in case of complex situation could help.
Still pondering on these presentations.