Polls convinced me Hillary Clinton wouldn’t lose. As an education researcher, the result was a wake-up call

Like a lot of Americans, I watched the election returns in shocked disbelief. All of the major polls had offered me confidence in a Hillary Clinton victory.

But while I was surprised by the results, I should not have been. I study school segregation, and my research has taught me one thing over and over: numbers alone are not enough.

To get an accurate understanding of situations, qualitative researchers like me believe that we must talk to humans. We must be out in the field forming respectful relationships, conducting ethnography and in-depth interviews. This effort to understand people’s lives and circumstances, interactions, beliefs, and social processes is important so that researchers — often ensconced with like-minded individuals and within liberal cities and universities — don’t miss what is actually happening on the ground.

I’ve seen the importance of this time and again in my own work. When conducting research for my book, “Public Housing and School Choice in a Gentrified City,” I wanted to understand why some groups opted out of their neighborhood school while others did not. To do that, I had to be deeply embedded in the community, talking to people about their complicated feelings about race and class and schooling.

If I had conducted a survey with questions about parents’ desire for “diversity” in their children’s schools, the results would have been misleading at best, since people use that word in so many different ways.

As I work to understand school segregation, I’ve also seen how the constant reliance on test scores to describe schools can alter the schools themselves and affect people’s choices. I have found that while advantaged parents in gentrifying communities would like to send their children to schools with “good” test scores, they also do not want their children in schools that are explicitly test-centered. This puts administrators of schools that must improve test scores in a bind and can reduce opportunities for integration.

Yet in education, we continue to place great value in numbers to judge the quality of schools, teachers, and student learning. We discount other factors and make decisions about teacher tenure, merit pay, the creation of charter schools, and school closings based on these flawed measures. Parents leave cities, buy real estate in certain communities, and select schools based on numerical school rankings.

We do that because numbers are powerful in their ability to convince us that they represent the truth.

That’s what happened to me this election cycle, when the poll numbers lulled me into missing the whole story. My parents had reported to me that wealthy, well-educated friends from their country club in the swing-state of Ohio, where I grew up, were Trump supporters. I heard about large parts of North Carolina, Ohio, and Pennsylvania with nothing to see but Trump yard signs; of Catholic-educated, one-issue voters from my childhood parish choosing Trump; and of a friend whose black grandson and his roommates were for Trump, too.

I paid this no attention. The numbers were very clear.

I don’t share these anecdotes to imply that they were valid research findings. But they do remind me how much qualitative research, done in a systematic, neutral way, might have contributed to the conversation before the election.

Pollsters, pundits, and I placed too much value on numbers alone while discounting the nuanced feelings on the ground in parts of middle America. We failed to explore this dynamic and how it could result in a Trump victory.

We can’t lose sight of this truth as it connects to education research. We must do our best to create new ways to understand school, teacher, and student success that are not oversimplified or misleading, and then advocate for their use.

We also need to support the work that gets us there. In education research, grants are far more lucrative and jobs far more plentiful for those who use quantitative methods. That data, when used well, has real value. But qualitative research cannot be overlooked.

Qualitative data is often wrongly disparaged for its alleged subjectivity, lack of reliability, or its cost. I would argue that the over-reliance on quantitative data in this election proved incredibly costly. Let’s not make the same mistake in education.

Molly Vollman Makris is assistant professor of urban studies at CUNY’s Guttman Community College. She holds a doctorate in urban systems and is the author of “Public Housing and School Choice in a Gentrified City: Youth Experiences of Uneven Opportunity.”

About our First Person series:

First Person is where Chalkbeat features personal essays by educators, students, parents, and others trying to improve public education. Read our submission guidelines here.