I also want to add these four main principles when it comes to standardized testing: #1: Standardized tests, if designed well, are decent indicators of how well your entire system is doing. Unfortunately most standardized tests are not designed well. #2: However, standardized tests in a vacuum are not particularly useful for telling you how the teaching in any particular school is going. There are far too many other factors to student achievement outside of the school that have to be normalized before you can even know what the standardized test scores are telling you. #3: Standardized tests, in a vacuum, are decent but not perfect at predicting how students are doing in academic activities. However, there's so much variation (great test-takers who fail at their schoolwork and great students who don't take the tests well) that they don't offer much information that isn't already available but looking at the students' body of work. #4: Teaching to standardized tests is useless when it comes to improving student outcomes. At best, you will get them to perform better on the test, but you will not help them perform better in school. In most cases teaching to the test will actually lead to worse school performance, as classes focused on test-taking skills are boring as fukk and don't help at all with motivation or other root issues in learning. Teaching to the test improves test performance for kids who are already well-prepared otherwise to succeed. Thus for the most part, standardized tests should really only be used to measure systematic issues. High-stakes testing to grade students for promotion/acceptance into higher learning or to grade teachers and schools on their performance is stupid. And teaching to the tests are stupid, because at best you'll only manage to mask your shortcomings, and more likely you're actually make your problems worse. I wanna hit that up in a lot of detail because it's 7-8 factors working together, but unfortunately I have to rush off right now. The issue is that the factors are pretty obvious, but most studies only look at 1-2 of the factors at a time, so they try to say "But look, remove this factor and there's still a gap!" because they're not recognizing the intersection of all the factors.