Why teachers, parents need data science

For many parents and educators, “data” is a four-letter word.
Parents have rightly begun to insist that schools prioritize data security to protect sensitive information about their children. Many teachers feel that the education data revolution is nothing more than a hammer that the world uses to beat them over the head. Every time a new set of test results is released, certain observers rush to interpret the results as a referendum on the quality of America’s educational professionals.
Debates about data security and accountability are important, and they have received a corresponding level of attention. I worry that this attention has obscured the reason some of us educators wanted to do this whole “data” thing in the first place.
Data is good. Data is powerful. Data can and should help kids.
Ask any parent or educator what they need to help kids. Inevitably, part of their answer is that they need more information. Why can’t Johnny read? Why is Tamara so far ahead of the other kids? Why is Raul acting out so much? At their heart, these are information questions, and these are exactly the type of questions that data science is designed to answer.
Data Versus Data Science
What is data science? Data science is a new buzzword for an old phenomenon: people who understand advanced statistical techniques and can employ them to answer questions of critical importance to particular economic sectors.
Today’s schools know the importance of data but have yet to reap the rewards of data science. Schools collect a ton of data on every aspect of their organization, but it is static. Once collected, data just sits in some central district computer. Sure, school districts send their data onto state and federal officials, where it largely just…sits some more. The state and federal governments will issue their reports, the vast majority of which educators never even know about and are not designed to help improve professional practice. The only real data-driven answers educators can get come from whatever standard dashboards come with the various computer systems their district uses.
Reports and dashboard metrics have their place, but the questions that teachers and administrators need answered change every day. To really help kids, education needs to shift the culture of education data towards greater flexibility and responsiveness to the needs of on-the-ground educators. It needs to build in next-level data use into the everyday experience of every teacher.  In short, it needs data science.
What Data Science Can Do For Kids
To start, every school district needs to understand that data analysis is as important as data collection, storage, and reporting. Education officials have access to a staggering amount of information about each kid. Data science can create links between all the different computer systems to answer questions that teachers and principals bring, questions that will help these professionals serve children better. In this way, data science can allow teachers, parents, and principals to get the information they think they need to do their job.
For example, let’s say Raul’s teacher approached her district’s “data help desk” looking for answers about his sudden bad behavior. The district’s data scientist might have already created a query that links the district’s multiple databases to flag known early warning indicators, so she is able to quickly tell the teacher that Raul’s increased acting out coincided with frequent bus delays that got him to school too late to have his breakfast. Just like that, the teacher knows to make sure to ask the cafeteria staff to save some breakfast for Raul and allow him to eat it when he arrives. Once he starts getting breakfast again, Raul again becomes a model student.
Imagine if every district had a data help desk, where data scientists worked to answer the information questions that came from real, live teachers. Imagine that every teacher had data-driven answers to the specific challenges they faced in their classrooms. Imagine if they got a weekly report that identified students who had yet to fall behind but who displayed early warning signs.
And early warning indicators are only one way data science can make education more effective. Data science can also help a teacher understand why certain students succeed, whether her new lesson plan was effective, whether her class really does lose focus after recess, and on and on and on. A data scientist is limited in his analysis only by access to accurate electronic records. Where those records exists, a good data scientist is sure to produce something that can help.
Data science is not a magic bullet. A school district with the most robust data science program possible will still face the countless challenges that seem inherent in modern public education, and overcoming those challenges will require the skills of capable professionals. A stronger data science program will make sure such professionals have the tools they need to succeed.

Elizabeth Sobka