Chicago Teachers Strike and Standardized Tests

Both the media and the critics of the current Chicago teachers strike are paying little attention to the primary reason that the teachers are not in their classrooms: Recent legislation passed by the Illinois legislature requires that scores from standardized tests be used for teacher accountability and performance.

To someone not in the education field, this mandate may seem reasonable. However, most people don’t realize that standardized tests were never designed to measure learning. Standardized tests were designed so that half of the test takers will fall below the 50 percent line. If 50 percent of students “fail” by design, how can these types of tests be justified for teacher accountability? Under this ruling, teachers will be judged and evaluated on a measure that is not valid for this purpose. Additionally, standardized tests are not reliable because they oftentimes do not get consistent results. (The same students taking the same test score differently.)

Realize, too, that many bright people simply don’t do well on tests. I know several respected adults who rarely demonstrate their competence, wisdom, or skill on a measurement instrument. The same goes for many students.

The problem is that too often educators and legislators attempt to make the social sciences have the same precision as the natural sciences. My post graduate experience as an economics major demonstrated this attempt. Economics is not a science—nor can it ever be. In contrast to the natural sciences, every measurement instrument in the social sciences is SUBJECTIVE. Yet we succumb to the illusion that only that which can be quantified can be true or valid.

Let’s be clear: I am not against accountability. I am against those instruments that are not valid or reliable, and this is the case of using standardized tests to measure teacher effectiveness especially when a teacher’s influence becomes known more in the future than in the present.

In short, using standardized tests to measure local teacher performance is invalid for at least three reasons:
(1) Standardized tests do not match their questions with the local curriculum or what has been taught,
(2) Standardized tests measure much of what students know when they first entered the classroom—especially in low vs. high socioeconomic areas—rather than what has been taught, and
(3) If the majority of students have learned what has been taught, then that information will not appear on a standardized test because it does not discriminate enough.

Standardized tests, by their very nature, are designed to discriminate, not to assess learning. As W. Edwards Deming (the international expert who brought both improved quality and lower costs to the workplace) said, “THE MOST IMPORTANT INFORMATION CANNOT BE MEASURED; THE VARIABLES ARE TOO GREAT.”

We should listen to the cowboy who said, “When the horse is dead, dismount.” Educators and legislators should take this wisdom as it applies to using standardized tests for learning accountability and for teacher evaluation. We should be using norm or criterion referenced measurements instead.