How Oklahoma’s A-F grading system discriminates against high-poverty schools — and how to fix it

by | December 10th, 2013 | Posted in Blog, Education | Comments (0)
Print Friendly

student2The debate over Oklahoma’s A-F grading system has in large part centered on a debate over the effects of poverty on student performance. Whatever you think about how much we can expect schools to overcome the effects of poverty, one point is not in dispute — that children living in poverty start from behind. On average, they have greater stresses and fewer learning opportunities at home before they ever enter the school doors. 

With low-income children making up 61 percent of all the students in the state, this achievement gap is a serious problem. That’s why we should be very interested in identifying the schools that are best helping these children improve. Unfortunately, Oklahoma’s current A-F system does the opposite. The method Oklahoma uses to come up with grades actually hides achievement gains among the lowest performing students while providing large bonuses to schools with students that start ahead. We can see why by looking at the three major components of A-F grades:

Overall Student Performance

First, all tests scores are placed in one of four categories: Unsatisfactory, Limited Knowledge, Proficient, or Advanced. From there, the “student performance” score calculation is relatively simple — it’s the percentage of “Proficient” or “Advanced” scores on reading, math, science, and US history tests. If a school has 75 proficient or advanced scores for every 100 tests, it will receive 75 points. This score makes up half of the points that go into the final grade.

This component of the grading system measures performance without considering at what levels students began or how much they may have improved or fallen behind over the school year. Even then, it’s an important metric to know. It helps us to see where achievements gaps exist and to set a goal for all students to become proficient.

Overall Student Growth

While it’s valid to evaluate overall performance, we should also incorporate student growth from year to year to be fair to those schools and districts that serve students who begin at lower levels. Unfortunately, the method Oklahoma uses to score student growth for A-F grades is very problematic.

The overall student growth score accounts for one quarter of school’s final score. This metric applies only to reading and math tests. After comparing two test results from one year to the next, it awards points in these cases:

  1. The student scores either “Proficient” or “Advanced” for  both years; or
  2. The student scores improve by at least one category from one year to the next (e.g. “Unsatisfactory” to “Limited” or “Limited” to “Proficient”; or
  3. The student scores “Unsatisfactory” or “Limited  Knowledge” on both exams but still improves by a number of points greater than the state average improvement.

The first metric of scoring points makes little sense. It awards a “growth” point even if the student’s test score declined. It awards a point if they scored “Advanced” last year but fell to “Proficient” this year. It awards a point if they scored “Proficient” both years but still had a lower score within that category.

Meanwhile, students in the “Unsatisfactory” or “Limited Knowledge” categories could improve without earning a point, if their scores grew by less than the state average and less than was needed to jump to a higher category. A school in which every student improved while remaining at “Unsatisfactory” or “Limited Knowledge” levels could easily earn less points than a school where every student scored worse or the same but began at higher levels. This adds even more bias towards schools that have most kids performing at grade level from day one.

This four-category method also makes the result too dependent on where Department of Education staffers set the cut scores. The Department already stirred up controversy earlier this year when they hiked cut scores for Biology exams after the tests had already been taken. How can we draw meaningful conclusions from A-F grades when the basis of the measurement can change arbitrarily from one year to the next?

Bottom 25% Student Growth

The third component of the grade is calculated in the same way as “Overall Student Growth”, but only for the lowest scoring 25 percent of students within a school. It accounts for one quarter of the school’s final score. This component continues all of the problems of the “Overall Student Growth” category and also adds another wrinkle. Putting extra emphasis on the bottom scorers within each school, rather than within the system as a whole, again slants the grading system in favor of schools with few poor kids.

The lowest performing kids in a high poverty area may be homeless. They may not have enough food to eat. They might be exposed to crime and violence. They are most likely to be dealing with major stresses at home that are beyond what schools can easily address. In contrast, students from a low-poverty area who struggle in school might have a learning disability that needs extra attention, but their problems are less likely to originate outside the school.

Of course a great many poor kids have safe homes and supportive families, and many wealthier kids may be abused at home or face other serious stresses. But on average, poverty creates a harsher environment outside the school walls. The bottom 25 percent faces very different challenges from one school to another.

How We Can Do Better

First, we should do away with the cumulative school grade. Kids don’t come home with a single grade at the end of the year. They may have an A in English, a B in math, and a D in science. Likewise, the A-F report cards break out grades between subjects, overall school performance, and student growth. Combining those into one grade based on an arbitrarily weighted formula sacrifices too much clarity for the sake of false simplicity.

We should also add grades for some of the important roles of schools that aren’t captured by standardized testing. Many of these metrics could be obtained without adding a large administrative burden on schools or a need for more testing. A few examples: What percentage of students participate in extracurricular activities? What percentage of parents volunteer at school functions or attend parent-teacher meetings? How many teachers have advanced degrees? What is the ratio of teachers, librarians, college counselors, and social workers per student? Is the school facility adequate — are there science and technology labs, well-maintained athletics fields, enough textbooks and classrooms, and a working heating and cooling system? What is the teacher turnover rate? What is the turnout for school board elections? These are just a few examples of more creative measures that would be valuable to parents and policymakers and could compensate for the limitations of standardized tests.

Second, we should award student growth points based on actual growth in scores rather than where schools fall within an arbitrary and changeable set of categories. The Department already calculates a scale score for each student, called the Oklahoma Performance Index (OPI), and uses it to award points for growth within a category. Applying that method to all students and eliminating the categories would be a fairer and simpler way to calculate growth.

Finally, we should improve how we assess growth of student groups that face special challenges. The lowest-scoring 25 percent of every school is not a meaningful category, because who those students are and what they need could differ dramatically between schools. Instead we should examine student growth in more meaningful categories such as low-income students, English Language Learners, or students in a special education classroom. Those scores would reveal what is working for the kids who need it most and help us to expand on those winning models. They would also fix the problem that the current system may be hiding progress for low-income and minority students, which puts our state’s No Child Left Behind waiver at risk.

Together these changes would make A-F grades simpler to calculate and understand, more informative to parents, and fairer to schools. We can resolve the grading system’s worst problems while still ensuring school accountability. Hopefully our state leaders will not let political posturing get in the way of a win-win solution for everyone.

Leave a Comment

You must be logged in to post a comment.