Bringing up big data and learning analytics typically sparks up the discussion. We have people who are in favor of it, believing that it is an essential component in improving future learning results. And then we have the opposing camp, fearing it will make teachers obsolete, throw privacy out the window, and in general make the world a miserable place.
I am drawing caricatures here, but this is still roughly the situation. I belong in the optimism camp myself, with certain boundary conditions: For example, the minute data travels beyond the classroom boundaries, it should be anonymized. Also, teachers should be in control when utilizing learning analytics.
Yesterday I encountered yet another list of big data challenges, shared by the TeachThought blog. The list was borrowed from another web source, namely OpenColleges. The list is as follows, and the items should be addressed in order to successfully implement big data in education:
- Value to the learner
- Depth of measurement
- Many factors influence learning
- Presenting data
For me these are self-evident, however quite different from each other. Allow me to group them into three main categories in order for you to see what’s going on. First of all, there is the “Understanding” category. That underlies everything else. We should understand what this big data thing is and what are its benefits. The following items belong in this category:
- Value to the learner, this is where it all starts
- Readiness to work with massive amounts of data
- Many factors influencing learning
- Presenting the data so that it makes sense
- Depth of measurement, going beyond traditional tests
When the above points are understood, it is possible to take the next step towards implementing the data-intensive learning system. Then the normal objections emerge. I group them into “Baseline” category:
- Transparency, knowing where the data goes and who is using it
- Openness on where and how the data should be utilized
- Privacy and delicate processing on personal data
Finally, there are some other challenges to solve. These are not really bid data -specific but concern any new technological innovations. I collect these into the “External” category:
- This is expensive. Yes it probably is, it should be considered as an investment
- This needs new infrastructure. Again yes, but gladly nowadays cloud services are there to help us
Big data by itself is no silver bullet. If you collect data but don’t have idea what to do with it, it is useless. You should first explicate your goals and then harness the data collection, enrichment, analysis, and reporting tools to drive you towards those goals. Then you implement it in a way that doesn’t violate people’s opinions on privacy, transparency, etc.
Flickr image CC credits: Kevin Krejci