Wednesday, April 8, 2009

The problem with standardized tests

Just posting something with this title would put me under pressure from the "reform minded" educators as someone who is against accountability, someone who thinks its okay if teachers don't do their job and our students don't learn. Thats not who I am. I do believe in assessing students' learning, and using it as a way to assess teaching. But I mean this in the most general sense.

Assessing students learning - good for students, teachers, districts, states, nations, to know how they are doing. Knowing how you are doing IS important as a student, it IS important to know how students in your class are doing, it IS important to know how students in your district is doing.

The problem is assessing accurately. I would argue that standardized tests aren't accurate (or precise for that matter). The problems with standardized tests are they are culturally biased, simplistic, and that you can do well on a standardized test without knowing content because of skills independent of the content.

First, the cultural bias leads to inaccurate results in a specific direction. Although there has been some attempt recently (at least reported attempt) of making tests reflect the full spectrum of american culture, there are still problems. Last fall, I was a human reader for the High School Assessment (HSA), a graduation requirement in MD. I read a story about someone canoeing through some river. One of the words in the story was aft, it didn't have any context clues about what aft might mean.

There was a question that the student I was reading to got wrong because she didn't know what aft meant. I consider aft to be a trivial word, not something that needs to be tested to earn a high school diploma. Not to say that students shouldn't know the word aft, but that there are literally a million words like 'aft' and that you can only learn these words through experiences and not by memorizing them from word walls or a handout list.

Second, tests are simplistic. They can't assess higher levels of Bloom's Taxonomy, only your ability to identify other people's analysis or criticism or evaluation. In a multiple choice test, I can't ask you to compare two works of fiction, only to identify which comparison the test maker came up with is the most accurate. The test maker can't assess someone's ability to do perform science, but only assess their knowledge of natural history or knowledge of science lab procedures. Only assessing a students ability to identify (never recall, and rarely create) the correct answer is simplistic, and the least we should expect of our students not the most.

Lastly, there are specific test taking skills that span content and are minimally useful to people in their every day lives. There can be two students that know the same amount of science content, that would be able to perform a task in a science lab as competently as each other, and who would be able to draw conclusions from their results as accurately as each other, but if one excels at taking tests (for whatever reason) they will not be equally likely to graduate. This for me is one of the worst problems. Its like building a huge mound in front of your house to keep the flood waters out, only to watch your home get flooded from your neighbor's yard. There is no reason to focus solely on content if the student would be better served by learning test taking strategies (that will be of little worth for the rest of their life). There is no reason to try and teach students words like 'aft' and have them have all of the rich experiences that is life, and that makes life exciting, and that teaches vocabulary, if you could just teach them smart process of elimination on multiple choice tests.

No comments:

Post a Comment