As much as I love testing, I can't help feeling that the state testing programs -- on which each state spends millions of dollars -- do not really help the students. It's not that they aren't designed well, nor is it that they are too difficult or too long. The reason they do little to help students is the turn-around time. The test is taken once a year, at the end of the year, and the results do not come until after the students start school the next year. So the turn-around for the test takes a whole year, during which time, the student may or may not get specific feedback and help concerning a test they took a year ago.
For the most part, students know when a test matters and when it doesn't. And the student test scores show this. For example, in Texas, the ninth and tenth grade scores are usually pretty low, but then all of the sudden, the students get smart and the scores go way up when they get in eleventh grade. It so happens in Texas that if they do not pass the eleventh grade test, the students do not graduate. Motivation does wonders for test scores.
School districts have gotten savvy to the game of testing and figure that giving tests that prep students for the big state test will help students do better. Districts have instituted what they call "benchmark" tests to determine student preparedness for the state test. These are often administered once a quarter, or in some cases, monthly. Like the state test, they are not graded and students know that doing well or poorly on the test does not affect their standing in the class. Except for some school districts that target students with interventions based on their benchmark scores, not many changes occur because of the benchmarks. How useful is this?
There are useful tests. They are the ones that teachers make, and are beneficial for students -- not just teachers and administrators. They are called curriculum-based (CBA) tests and they are what teachers should be teaching to.
Are the CBA's perfect? Not hardly. So that is where all of the teacher effort, principal effort and district curriculum effort should go. Rather than spending valuable time on preparing for a minimum standard state or benchmark test, teachers should focus on getting students ready to pass a CBA test.
The Nuts and Bolts of CBA
The ideal teacher test is designed before instruction begins (according to McTighe & Wiggins in Understanding by Design). Each question is correlated to a specific and prioritized student-learning objective, and is designed to be either be easy, medium, or difficult to differentiate the test for all students. The students are given the pre-test before instruction begins, and if all of the students pass with 80 percent or better, then the teacher can compact the curriculum to be taught and then move on to the next unit more quickly.
But, if full instruction is necessary, according to pre-test scores, then the value that the teacher added will be reflected in the post-test scores minus the pre-test scores. Additionally, since both the pre- and the post- test are correlated to the student learning standards, then in rapid fashion, a teacher can identify specific student learning needs in a quick turn-around time, and then re-teach them in a better, more effective way.
This type of ideal test can be time consuming, even though the benefits outweigh the hours spent in creation. One way to mitigate this concern is to allow the ideal tests to be created by a team of teachers, all agreeing to a specific standard for the correlation and the quality of each test item. This method has the effect of holding each of the other test writers accountable for good instruction in their perspective classrooms. It also raises the quality of the tests because more eyes are critically analyzing each test for errors in content and format.
Education author and professor Fenwick English champions that the test that should be "taught to" is the correctly designed and correlated CBA and not the minimum standard state test.
How do you make useful and beneficial CBA's for student?