With Big Data Comes Big Opportunities and (Maybe) Big Risk

We live in an age awash in data. Every age has spawned data of various forms, but what sets today apart from yesterday is the fact that impressive computing power and abundant and cheap digital memory are combining to inform behavioral analysis and prediction in impressive ways. Collating millions of bits of academic and economic data would have taken decades, maybe centuries, only 50 years ago. Now, number crunching happens in seconds, and data storage is cheap and forever.So, it was only a matter of time before the quants turned their inferential statistics on student performance and began using big data to predict student success in higher education. And there is no reason why K-12 should be exempt from the trend. As Time Magazine points out in a December 10, 2014, blog by Jon Marcus, an increasing number of American universities are using data algorithms to predict student success, and direct interventions toward those most at risk for dropping out.At least that is the putative rationale given for using data in this way. It doesn't take much imagination to see the less lofty and humanistic possibilities inherent in big academic data. At what point does identifying students at risk become screening them out from admission in the first place? And what if, as I suspect will be the case, certain subgroups of individuals seem more prone to drop out?Big data can be a boon to health care, leading to better and more useful therapies. But it could also be an asset to insurers, seeking to limit their liability exposure by keeping high risk cases out of their risk pool. Data in any field is not an unmitigated good; rather, it is in the application and unintended consequences that the good and evil emerge. Proponents of quantitative methods in educational forecasting would do well to remember Google's original dictum, "don't be evil." Seems quaint now, doesn't it?