Data, as anyone who utilizes such online giants as Netflix and Amazon can attest, is being used in an increasingly number of predictive uses. As advanced algorithms compare data points for shopping or viewing trends then use these trends to successfully predict likely trajectories of online viewing and purchases, it seems only natural that other fields adopt similar technologies.
This past weekend while attending the Big Data Summit in Denver sponsored by Tech & Learning and Bright Bytes, I was able to experience how this predictive analytic approach based on large amounts of data points could positively influence the future of education, altering the trajectory of not only students, but the profession itself.
For approximately the past ten years, data-based decision-making has been a common buzz word in education. While grounded in the best of intentions of removing subjectivity and biases from educational placement and intervention decisions, any researcher understands that the effectiveness of this approach to making decisions depends heavily upon the amount and validity of the date being analyzed. By compiling benchmarking scores obtained sometimes in as little as fourteen minutes, data-driven approaches in education often focus on very precise areas (i.e. math computation, or reading phonemic awareness) that can then be addressed with a specific math or reading strategy backed by research, a very successful scientific approach referred to as RtI (Response to Intervention.)
With the advent of promising new approaches to data collection, education has the opportunity to address early in a child’s academic career certain risk factors that have been correlated (not necessarily in a causal relationship) to disengagement and eventually dropping out of school. After witnessing Dr. Kristal Ayers’ @kristalayres1 presentation on the Diary of a Teenage Dropout which highlighted the scientific gathering of many various types of data points (attendance, family structure, tardies, grades, disciplinary infractions) IN ADDITION TO more traditional data points, including test scores, can be inputted over time and used in conjunction with an algorithm to determine students in “Early Warning” phases, even as young as grade school, it became evident to me that what the education field has long been referring to as “data-based” decision making is exactly that: based on data. What Bright Bytes and Dr. Ayres are pioneering is far more innovative: it can be said to be data-infused decision making, for it is an approach inundated with nothing but vast amounts of (seemingly) unrelated data surrounding students.
While such an approach shows immense potential in curbing drop-out rates and decreasing the number of students failing school across the board, its true value will be measured in a set of data much more difficult to quantify yet vitally more important: preventing schools from failing students!