Guest Post: Your Data Will Lie to You, if You Let It

By Eric (Big “E”) Mihelbergel: Eric Mihelbergel resides outside of Buffalo, NY with his wife and two daughters. His education advocacy, which started in his hometown of Tonawanda, has spread across the entire state of New York.  Mihelbergel is a founding member of the groups New York State Allies for Public Education and Western New Yorkers for Public Education. He is an active member of the Opt Out movement and administers the Facebook page NYS Refuse the Tests. Along with Leonie Haimson from New York City, Eric and I will be presenting at this upcoming forum in November.

 * * *

One of the great flaws of human nature is that our ego makes us suddenly believe that things magically become more important if we measure them. Data does not exist naturally in the universe. It is a human invention, and thus it is based on human assumptions. Data itself has no power. But all too often humans allow data to tell lies.

Consider the case in the Ken-Ton School District where my two daughters attend school. The school district recently mandated that teachers use NYS Common Core modules in grades 6-8 ELA and Math because there was a decline in NYS test scores. Although they are compassionate people, they poorly assumed that test scores are down because the wrong teaching methods are being used. But the data does not say this at all. The data only says that test scores went down, but nothing more. After data is collected it is then the job of humans to interpret the data based on human assumptions, but if the wrong assumptions are made then the data can cause great harm even if it was meant to help.

In this case, the district assumed that there were no other variables. In fact, however, there are many variables.

  1. Over the last year there has been an influx of lower income families from the city of Buffalo into the Ken-Ton school district. Research shows a direct relationship between low income and low test scores.
  2. A significant number of high performing eighth grade students were not required to take the NYS tests this past year because, for the first time, they could opt to take only the Regents exam. Obviously, if high performing students are not taking the NYS test then the overall score will be lower.
  3. There was a large increase in the number of test boycotts in the district this past which could affect overall scores depending on the demographics of those students.
  4. One of the middle schools had major construction happening outside the windows of testing rooms which can pose a great distraction to students. If students are distracted they will likely score lower.
  5. These tests only test roughly 15-20% of the standards which may not be what is most culturally important to students and families in a particular community, therefore they may not focus on that particular 15-20% of the standards in their life. Who are we to change that?
  6. The entire county had lower scores on these tests indicating that perhaps the test as a whole was more difficult than the previous year.
  7. Over the last year there has been an influx of refugee families into the school district. Many of these families speak English poorly and have little formal education, so certainly they would score lower on tests.
  8. There have been many cases across the state where students purposely score poorly to hurt teachers and principals that they don’t like.

As you can clearly see, there are so many variables that it would be asinine to assume that test scores were lower simply because of teaching methods. No one can logically buy that. By making a poor assumption, the district has allowed the data to tell a lie which has now harmed students by mandating a curriculum that takes control away from the teacher. The teacher has now lost the ability to choose a curriculum that can connect with students in a way that is best tailored to their needs because the teachers are now forced to use a cookie-cutter curriculum designed by someone far away that has never met these children. The exact same data that the district thought would help students is actually hurting students. Data can tell very severe lies if not handled with great care.

So what is the better alternative to data-driven education? The answer is teacher-driven education, not data-driven. Let’s look at the numbered variables above from a teacher point of view.

  1. The teacher has direct contact with these students and would be able to personally work with the student to truly know the student’s abilities, learning style, fears, passions, distractors, focus level, and personality.
  2. The teacher can differentiate between students that were taking Regents exams and students that were taking NYS tests.
  3. The teacher teaches all students regardless of whether or not they have taken a test.
  4. The teacher guides the child back into the learning environment.
  5. The teacher learns what is most important to families and communities and provides an education that is best suited to them.
  6. The teacher assesses each student every day through personal interaction so that students are fairly assessed.
  7. The teacher can quickly identify these student’s needs and make minute to minute adjustments to help them progress.
  8. The teacher knows the students well enough to be able to confront this and guide the students back to focus.

Data can do none of this because data is simply one small tape measure in a very large tool shed.

So why were these poor assumptions made that are ultimately harming students? I have no reason to believe that there is any malicious intent in interpreting data in a way that harms students. The decision makers are very compassionate people. More likely, it is a result of the personal beliefs and assumptions of the individuals that viewed the data and mandated the NYS modules. Their background, history, and paradigm shows them to be strong believers in the Common Core, NYS modules, and data collection. Therefore, when the data was viewed, their brain very likely filtered the data to make it seem like the data supported their personal beliefs and assumptions. Again, there was likely no malicious intent at all, but the human brain framed the data to fit a paradigm. This is the danger when data is filtered through a biased human mind. The data becomes an unexpected danger to our children.

2 Comments Guest Post: Your Data Will Lie to You, if You Let It

  1. Rastamick October 22, 2014 at 4:36 pm

    Great piece, Eric. This is one to keep handy for shutting up the enemy with simple fact. Thanks.

Leave A Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.