Not an “either-or” scenario

Despite the many potential benefits that come from an overlap of the fields of data science and education, Loretta “Lucky” Mason-Williams is very clear that there are and will always be things a teacher can tell you that a data set cannot.

“Did a particular student have breakfast this morning? Is this student behind because we haven’t gotten to that part of the curriculum yet?” asks the associate professor of special education in the Department of Teaching, Learning and Educational Leadership (TLEL).

While an increasing amount of data is available, Mason-Williams is confident that the need for a teacher’s intuition will always remain constant.

“Having more data allows you to make more concrete decisions by providing more reliable information. The disadvantage is that people may forget that there is always more to the story, and you always need to have a teacher there to connect the dots,” she says.

With an optimistic outlook on the future of education, Mason-Williams sees an opportunity to significantly strengthen a teacher’s intuition by backing it up with powerful data.

That’s why she’s immersed herself in multiple data-heavy research projects and currently serves as a member of the steering committee for Binghamton University’s Data Science Transdisciplinary Area of Excellence (see sidebar). It may be surprising to learn that her love of data is a fairly recent development.

“I began my doctoral work hating quantitative research and the math involved with statistics,” she says.

It wasn’t until she began digging into research that utilized large data sets to understand and unpack issues of education equity that she realized the full potential data science has for her field.

“It allows me to understand how this information can help us think differently about how we’re providing resources to schools,” she says.

For the past decade, her research has focused on exploring inequities in special education across the United States.

According to Mason-Williams, the special education field has a significant problem with teacher retention, especially in areas where working conditions are difficult. Because of this, she focused on these areas to learn where improvements could be made.

“The goal is to provide better, more-accurate data on who’s teaching students with disabilities, as well as what policies could be put in place to better recruit and retain teachers,” she says.

Mason-Williams says her research so far has mostly relied on big data consisting of secondary sources and teacher surveys. This has allowed her to understand the qualifications teachers are bringing to the classroom, as well as how some areas have more qualified teachers than others.

For instance, she identified that teachers and principals in specialized, alternative schools specifically for students with substantial learning and behavioral needs may not be as qualified in special education or in content-area instruction as their colleagues in traditional schools. She’s now looking to incorporate big data into her research to discover exactly why that is.

“I’ve been working on a couple different fronts to get state-level data sets that fall under this idea of big data. This would be population-based data that’d be coming in fast and furious, and would include data about teachers and students within those areas,” she says.

The combined examination of these two kinds of data would allow Mason-Williams to develop more concrete findings on how state-level policies affect special-education quality. This would add another relevant layer of insight into her already robust research into education equity.

“For years, social scientists have relied on survey data or census data that was collected a few years prior. But now we can use data immediately, which allows us to forecast,” she says.

But with this powerful ability comes an equally powerful need for responsibility, especially when dealing with data that could affect the education of millions of students.

“There are some ethical issues with forecasting that I think need to be part of the discussion,” she says. “We’ve made a number of mistakes in education policy based on misrepresentations of one or two studies. That’s the scary part.”

And with ethics at the forefront, Mason-Williams is hopeful her work will make a positive impact.

“I’ve found in meetings with administrators and policymakers that they are anxious for us to have better answers and to better understand where and how to distribute teachers. The power of these bigger data sets is that we may not come up with the answer, but rather multiple answers,” she says. “It’s a multifaceted problem that needs multifaceted solutions.”

And despite the improvements that can be made, Mason-Williams does not hesitate to champion America’s education system.

“We have amazing teachers and dedicated leaders in our schools. We are so lucky to have our public school system in the United States, and we don’t herald it anywhere near as much as we should,” she says.

We don’t laud teachers as much as we should, and I’m excited about how big data can support making them better — not fixing them.

Loretta “Lucky” Mason-Williams, associate professor of special education in the Department of Teaching, Learning and Educational Leadership