Counting down the days...5

It's the weekend...five days left.

Last week I attended a retirement dinner for the four people from my school who are retiring. One thing that came out during the "entertainment" was that I don't particularly care for standardized tests.

While that's true, what was left out of the discussion was why I don't care for them. In my opinion, standardized tests have their place, however, that "place" has long since been removed from the educational landscape.

Standardized tests are just that - standardized. They tell us how students compare to others who have taken the test. This can give us information about what academic weaknesses a student may have compared to others. In other words, it can be diagnostic. Limitations of standardized tests have to be considered when using them to diagnose. There is information, however, that standardized tests supply which can be helpful.

Walt Gardner wrote about the standardized tests which teachers take to "prove" that they are good teachers...and about how the paper and pencil test does not really prove performance ability. The same is true for students. Standardized tests do not measure everything. They need to be used ONLY for what they do measure.

The issue of standardized tests being misused is further complicated by the push by the current administration to use student scores on standardized tests to evaluate teachers. Standardized tests do not cover everything learned in school. They weren't developed for that purpose. Further, there is no evidence that they are accurate measures of teacher effectiveness.

One of the most important things about science and statistics is that in order to make a conclusion about a hypothesis, one has to make sure that external variables are not influencing the outcome of an experiment. In other words, if you are hypothesizing that sunlight is necessary for plant growth, and you have designed an experiment in which some plants receive more sun than others, it's important to make sure that other variables like the amount of water, soil quality, and air temperature are kept constant. If they're not, then you cannot assume that a change in growth was due to light differences, and not something else.

The same is true for standardized tests. Teacher A works in a school in a wealthy district. Her students are well cared for. They receive regular dental checkups, adequate meals, plenty of rest and exposure to literacy rich environments at home and at school. Teacher B works in a school where 80% of the students live in poverty. Many of them only have two decent meals a day...the breakfast and lunch they get at school. They may have dental problems which have never been identified. They may have difficulty sleeping in crowded homes with little exposure to print. There may be absent parents. Their medical, nutritional, and emotional needs may not be adequately met.

It's clear that the variables outside of the school make things more difficult for the students of Teacher B to learn. Should both teachers be subject to evaluation based on student test scores?

Even if we use a "growth model" in which the amount of growth the students make during their time with a teacher is used to grade the teacher, the variables get in the way. Middle class and wealthy students will have fewer outside influences which interfere with their growth than children living in poverty.

Finally, much time is spent norming standardized tests to make sure that they accurately measure that which they claim to measure. There are, as yet, no methods for establishing whether standardized tests of students accurately measure teacher competence. There have been no studies done to establish the effectiveness of standardized tests for that purpose, and the tests themselves were not developed for that purpose.

During the presidential campaign Candidate Obama said that we shouldn't judge students based on one standardized test. Yet, the department of education is now encouraging school districts and states to rely on those tests to judge teachers and teacher quality.

Using inadequate and inappropriate tools to measure teacher quality will not help improve teaching and learning. We need to develop better and more complete ways to evaluate teachers. We need to develop better and more complete ways to evaluate students.