Over the course of my career as an educator I have developed my own data to drive my decisions on what needs to be emphasized in terms of classroom instruction, curriculum benchmarks, skills, etc. Simply stated, students in my eighth grade U.S. History classes take a pre & post test covering the major benchmarks of the school district curriculum (which correlates with the State of Michigan benchmarks as it should). Within the first two weeks of school all students take a pre-test; this is the same exact test that is given the last week of school as a "final" exam. This pre-test is corrected immediately (the following day) but the results are not entered as a grade and I ensure that students know this prior to the fact to eliminate anxiety, etc. Quite frankly, I tell them they should "flunk" it because its everything they are going to learn this school year. From the results of this pre-test I can determine areas that my class load of students are strong and weak on. As such I can then adjust my instruction accordingly for the year. The pre-tests are collected and kept until the end of the year and then returned to students after they take the "final" exam. Students are then able to compare and see how much they improved and learned throughout the school year; It's a great way to conclude a school year... on a positive note.
The data gathering from administering a pre & post test, as described previously, is the most telling and useful available (also much cheaper) to students, parents and educators. When it comes to making decision/s in education using data the most important factor is the usefulness to improving a students life, learning and skills. The structure or test used to gain data must be focused on the curriculum taught in the classroom and who better to know it than the classroom teacher. Testing is not about trickery, its about determining learning outcomes and the questions must reflect this very concept. The data from tests needs to be instantaneous; immediate feedback is necessary to the learning process of the individual. As such, I would suggest to everyone, when looking at educational data be careful jumping to conclusions about student success and teacher effectiveness from standardized tests because inherently these tests are flawed. The best indicator/s to show student success are what teachers correct, send home and communicate to parents.