Mayer -Schonberger, Viktor and Kenneth Cukier. Learning with Big Data: the Future of Education. Houghton Mifflin Harcourt, 2014. .
- Olgerta Tona, Ida Asadi Someh, Kaveh Mohajeri, Graeme Shanks, Michael Davern, Sven Carlsson and Miranda Kajtazi. "Towards Ethical Big Data Artifacts: A Conceptual Design." 2018. Accessed 2 February, 2018, http://engagement.hicss.org:81/papers/0573/paper0573.pdf
This work is important in its focus on how big data is profoundly shaping education today. Mayer-Schonberger and Cukier highlight the benefits, but also points out the risks and dangers. First, the benefits: “It is a lens into three core qualities that will improve learning: feedback, individualization and probabilistic predictions.” To appreciate the import of this comment, it is key to understand the difference between what we did with data in the ‘small data’ era and what we can do now. In the era of ‘small data, doing data analysis was costly and time-consuming. Limited amounts of data were collected because it was difficult. The objective was to answer (find the results) to a question already asked. Researchers were looking for causality. In the world of big data, (now that the costs of data storage and data access have plummeted), we are able to uncover correlational insights. The feedback in this case is about the process. Using masses of data enables probabilistic predictions and discovery of insights that were impossible to uncover before.
When scale changes from small to ‘gigantic’, the scale completely tips in favor of correlational insights over causal link. In relation to my research, this suggests the importance of having access to a one – information view of all permitted data access, and having access to big data tools that enable knowledge iterative discovery In terms of education, the Holocaust is a ‘big data’ topic. Being able to examine Holocaust-related information at scale, ope4ns up the possibility of teasing out new insights into the data that just cannot show up in small data amounts. An example of this kind of outcome is the initial data analysis from UT Dallas on “Digital Studies of Deportations.” Using GIS data analysis tools to comb large data sets of deportation numbers correlated with Geographic place data and movement of top Nazi officials in specific European countries during the last years of WWII, has already translated into new insights.
As far as education is concerned, Mayer-Schonberger and Cukier make very clear how big data is helping to drive the change from ‘formal – assembly line education’ to individualized, dynamic learning, that takes advantage of learning changes in real time, based on changing feedback from new kinds of data collection that can be fed back to both teachers and students. For the first time, an enabling technology – big data- can make adaptive learning possible. Schonberger see this technique as “leading to a new era of highly personalized instruction.”
As the authors put it:
not only will this {sic: big data} information be used to redesign what already exists, but it can be analyzed in real time, to automatically present materials that are the best fit for the student’s specific need at a particular moment.
These outcomes inform my research into the impact of artifacts on Holocaust remembrance because big data tools enable new visualizations and interpretation of information. Another big data source is computational reading. Applying computational reading to Holocaust literary artifacts opens up the potential for comparing across literary genres in new ways.
In summarizing the dangers of big data, Mayer-Schonberger and Cukier discuss three things: academic tracking of students, leading to a narrowing of their opportunities in life. Then privacy – “and not just the ‘conventional concern over privacy and data protection, but a more unique worry - “the threat of an unshakable past. Mayer-Schonberger and Cukier go on to note: “Big data doesn’t simply magnify these problems, it changes their very nature. Here, as elsewhere, the change in scale leads to a change in state.” Changes in ‘state’ indicate profound changes in the status quo of student educational outcomes and even career access for students in their adult years.
The authors do not say that the search for causality is wrong, or should be given up. They suggest that rather than hunting at great expense for an often elusive why, we may be better off with a more pragmatic approach, of aiming to first comprehend the what that non causal analysis can reveal.”
The challenge is that there is a power imbalance between individuals, organizations and society in the big data sphere. There are no solutions out there yet that adequately and ethically accommodate the conflicting interests of individuals and organizations. In fact, the technical solutions so far are woefully behind the ‘technical wizardry’ of big data value extraction.