Every day, Netflix collects millions of searches and clicks to tailor movie recommendations for subscribers. Hospitals crunch the numbers on medical statistics to predict patientsí likely needs and calculate risk. During the past election, President Barack Obamaís campaign scoured consumer information, voter contact reports and demographic data to target voters block by block. ďBig data,Ē it seems, is seeping into every area of our lives. But one holdout remains: the classroom.
Data may not be absent from schools for long, though. Two states ó Oregon and Delaware ó are trying to determine how big data can help improve teaching and learning. In a new report I wrote with Jennifer Cohen Kabaker, we highlight the most innovative practices of those states and flag potential obstacles as more states begin following their lead.
The gathering of student data began long before Googleís heyday. The passage of the No Child Left Behind in 2001 heralded a new age of student data. States, districts and schools were required to publicly report studentsí outcomes. That kicked off a decade of development in statewide data systems that could collect the kind of information Congress asked for: demographics, graduation rates and studentsí test scores, primarily.
For all the states collecting data, though, few have managed to produce much helpful, timely information for teachers. In many cases, student test scores arenít available to educators until the school year wraps up. Even fewer states have given teachers the skills they need to use student data.
Oregon and Delaware launched their new data programs with federal dollars. Those funds paid for data experts to teach educators how to analyze, understand and value data. They showed teachers how to read the statistics the state provided them and how to adjust their classroom teaching so that studentsí needs were met. An evaluation of the Oregon project showed that its teachersí student test scores improved and even surpassed the stateís other students. The evaluation also showed that in both states, teachers are more comfortable using analytical approaches to instruction.
As education data trailblazers, both Oregon and Delaware remain models ó and cautionary tales ó for the states that will follow their lead. Timely access to student information, more than just once at the end of the year, helps Oregon teachers identify which students donít yet understand new material, and allows Delaware teachers to see which students respond to different teaching styles. These are the kinds of details teachers canít intuitively know in a classroom full of kids all at different skill levels.
But the two states also illustrate how far many teachers will have to come. Only 2 percent of 180 teacher preparation programs surveyed by the National Council on Teacher Quality sufficiently covered data analysis skills, and only 16 states required data literacy as a condition of teachersí certification and teacher education program approval. And professional development programs are usually superficial, one-time efforts, so they canít often embed a culture of data in a school. If teachers donít even know how to read the information states collect, how can we expect them to adjust and improve from it?