Standardized Testing - Did You Know?!?!

I must start by giving credit to Gerald W. Bracey's article "Put to the Test" for some of this info.

Before "Obamacare" there was a different kind of big-time legislation called No Child Left Behind. NCLB basically said that every student in the country would be at grade-level by 2014. As a realist, I scoffed at this idea. As an educator who knows about the tests that will be used to prove this, I knew it was impossible. See, test-makers won't let everyone pass- EVER!

Let me explain. There are two basic types of standardized tests: norm-referenced (NRT) and criterion-referenced (CRT). NRTs give results in percentiles and are compared to grade-level. These test scores are basically lined up against one another and ranked. CRT results are given as a numerical score with cut-offs at specific numbers that tell students if they pass or fail. These tests are compared to a total possible. If we were to use NRTs, no more than 50% of students would "pass" because test-makers compare scores to a "norm" so that scores are distributed widely. Since it is comparative to other scores, not everyone could be considered "at grade-level".

 I think (but have no data to verify) that most states use CRTs. ISTEP+, Indiana's high-stakes standardized test is criterion-referenced. Aside from arguments about whether or not tests truly show student knowledge and skill, let's look just at the test-making process. Each question is sampled to see if the results are "valid". Here is a list of reasons a question would get omitted:

• a high percentage of students get the answer right
• a high percentage of students get the answer wrong
• students who got most test questions right all missed the same question, but students who didn't perform well overall got it right.
• students who bombed the test got the same question right at a high rate

Basically, if most students get a question right, it is removed from the test. ISTEP+ is supposed to be based on the Indiana Academic Standards which are publicly available and used by teachers to plan curriculum. If teachers are teaching the same skills and concepts, a large majority of students will become skilled at those skills. By my understanding, these questions would then be removed from the test or changed so that more students miss it. That could be achieved by changing the distractors (the wrong answer choices) or the wording of the question. This sounds to me that if kids get good at something, test-makers will try to simply trick them into wrong answers so that the results from the tests are more widely distributed.

 This is the process behind tests that can determine if students can continue to the next grade or even graduate. Scores from these tests will be used to evaluate teachers.