Big Data in Education
By President Susan Fuhrman, Teachers College, Columbia University
In a year marked by widespread concern about the extent of government data collection and mining, educators and parents have paid surprisingly little attention to the amount of data collected in schools. Conversations about data privacy generally focus on security and protection from breaches and hacks, but increasingly are addressing the multiple uses of student data and the awareness amongst students and their parents.
Imagine Katie, an 11th grader in a high school of about 600 students She has excellent attendance, gets good grades, performs slightly less well on standardized tests than her grades would predict, and is involved in multiple extracurricular activities Katie is interested in attending a liberal arts college and is hoping that the academic honors she has received will help her. All of Katie’s information is kept in a school information system, which she and her parents can log into in order to check her grades and standardized test scores. Several school entities use these databases to track students and compare performance across schools and teachers. Recently, a number of states and at least one private company, InBloom, are linking databases across levels of schooling, including same state college and postgraduate training.
Katie enjoys using the course management system for each of her major classes. She can check and submit assignments, enter discussion forums with classmates, see her grades and access additional resources. Her teachers and administrators use the learning or knowledge management system to organize and manage all class information and follow individual students throughout the year.
For physics, Katie uses educational software that allows her to simulate experiments. The software adapts to her input by giving her hints when she is having trouble solving a problem. This adaptive educational technology (AET) teaches Katie content and skills. It also records her answers and response times, the hints she uses, and the aspects of the software she attends to—providing insight into how Katie masters the material. AETs are not only useful to students as they learn and track their own learning, but are also valuable to teachers, who can modify or personalize instruction based on the information the software provides about where students are encountering difficulties and/or excelling.
New platforms are emerging that connect the three types of databases storing educational information about Katie. In addition to Katie, her parents, and her teachers, there are two additional prominent users—researchers and developers. Researchers frequently want access to data collected by various software platforms because they can study the progress of students in various contexts, link progress to student backgrounds and classroom characteristics, and assess the effectiveness of teachers and schools. Using data collected by learning management systems, they can study the impact of various curricular or pedagogical approaches.
Commercial developers who create the platforms and programs that collect data are often asked by educators to provide ‘analytics’ or to find patterns in the data about students, teachers, classes, schools, etc. Commercial entities may also want access to the data in order to refine and improve their products and to develop learning materials that are responsive to the needs of students and therefore of interest to buyers.
Although these users are given data that is coded to protect individual identities, it is not impossible for those identities to be revealed. A determined snoop could identify Katie through her unique course-taking, extracurricular, and academic record in a relatively small school.
Educators should be leading a wide-ranging conversation about how to protect the privacy of students. In a letter to the editors of the New York Times several months ago, I suggested that all the providers of data, especially students and parents, and all the data users should be addressing questions such as the following:
How secure is confidentiality when students’ names are coded to protect their identities?
Who owns and controls access to student data?
How do students and parents consent to the data collected about students? Are they informed about all the various possible uses for that data, and about security procedures?
More data can help educators improve learning, but at what cost to the personal aspect of interaction with students? Will the premium placed on what’s measurable depreciate teachers’ perception and judgment? Will recorded early missteps limit students’ ability to get a fresh start as they move on?
While recognizing the potential value of better educational data, educators and citizens cannot avoid wrestling with the hard questions about privacy, student-teacher relationships, and the infinite number of less measurable variables that bear on the education of every child.#