The Problem with Standardized Tests

Each state mandates that public school children take norm-referenced or standardized tests every year. Most refer to these tests as end-of-grade tests or benchmarks. Not every child in every grade takes one of these tests. When and if your child takes a standardized test depends on the child's grade and where you live.

Standardized Testing Scores

When a student takes a norm-referenced or standardized test, they receive a numeric score and a percentile score. The percentile score is not the same as a percentage score. A percentile score is a comparison of one student's performance on that particular test to the performance of other students in the same grade who took that test. Most students score in the 50 percentile range. A percentage score is a score you would receive on a math test taken in class. If your grade was 95 percent, then you know that you answered 95 percent of the questions correctly.

Think of the percentile scores as all of the test results for any specific grade plotted out onto a bell curve. The majority of the students will perform close to the middle of the bell curve which is an average score. Students who perform very well or those who perform poorly will be on the outside edges of the curve. Herein lies the problem with standardized tests. You cannot accurately compare the results of one brand of tests to another. For example, scores from the Iowa Test of Basic Skills are not comparable to the test scores from a TerraNova test.

The Problem

The "No Child Left Behind Act" mandates that public schools show adequate progress in student performance from one year to the next. Failure to meet the "No Child Left Behind" guidelines has a negative impact on the amount of funding a school system receives. Failure to meet the guidelines in successive years require sanctions be taken to improve the school. In order to demonstrate student progress and conform to guidelines; school districts use norm-referenced or standardized tests.

Scores from different norm-referenced or standardized tests are not comparable to each other. Every brand of test has its own calculations for determining scores and percentiles. In addition, not all tests cover the same subjects. Every state has their own rules concerning the type of test used and subject matter that needs to be tested. To complicate matters further, many states such as Connecticut, Texas, Florida and North Carolina have their own end of grade tests or benchmarks. Other states use tests such as CAT, TerraNova, Iowa Test of Basic Skills or Woodcott-Johnson.

Now throw homeschools into the mix. In many cases, homeschoolers do not take the same tests as public school children. In North Carolina, where standardized testing is required of all homeschooled students aged 7 through 16, homeschoolers are not permitted to take the North Carolina End of Grade Tests. Instead, they must make arrangements to purchase and administer a nationally available test like the Iowa Test of Basic Skills. In Arkansas, homeschoolers may take the mandated test offered by the state at no charge or they can arrange and pay for private testing. The homeschoolers in Arkansas do not take the same test as the public school children.

This leaves us with the question-how can we compare student performance across the country if each state is using a different test? The "No Child Left Behind Act" defines student performance. It includes special situations like ESL students and special needs students. The guidelines do not take into account the fact that each norm-referenced test evaluates student performance differently. Homeschools and schools that do not receive federal or state funds are exempt from "No Child Left Behind" guidelines.

The Solution

In order to truly compare student performance across the United States, a national test should be used. This makes measuring performance fair to all. All students should have to take the national test as set out by the educational code of each particular state. Using a national test will ensure that scores in the 50 percentile in Texas are comparable to scores in the 50 percentile in Maine. In states where homeschoolers are required to take standardized tests, the homeschoolers should take the national test.

Perhaps a better solution is to do away with standardized testing. Even with a national test, students with learning and physical disabilities will have to take modified versions of the tests. Usually tests taken with modifications are not included in the percentile scores. Eliminating the standardized test creates many problems. The biggest challenge is figuring out how to measure each school district's compliance with the "No Child Left Behind" guidelines without using some form of standardized testing.

Using standardized tests as a way of measuring student performance has created a situation where student progress in one state cannot be compared to student progress in another state, making enforcement of the "No Child Left Behind Act" arbitrary and unfair.