Arizona is lowering the passing scores on their state-level standardized tests, and apparently they're not the only ones:
Arizona isn't alone in lowering passing scores on standardized tests and setting up dual rating systems to help schools meet tough new student achievement goals. Many other states have chosen to drop the academic bar to give schools time, and room, for improvement over the next decade...
This year, Arizona will lower its proficiency rate for the math portion of Arizona's Instrument to Measure Standards, the big state test. The modified test will reflect what state Superintendent of Public Instruction Tom Horne calls a "more reasonable" expectation of what Arizona students can do.
This year, students must answer 20 out of 36 reading questions correctly and 21 math questions out of 40 correctly. By 2005, those numbers will jump to 24 in reading and 27 in math.
The article reports similar standard-setting fluctuations in Texas and Colorado.
Given my limited experience with standards setting, I can't say that I know the perfect way to do it, but I know enough to have a great deal of sympathy for the state organizations that are attempting to do it. Sometimes standards are set too high and must be lowered, but I think that doing so creates credibility problems. What if the standard is too low (and how low is too low)? And how can test takers assess the standard without information about the difficulty of the test, and assurance that the test difficulty will remain constant across the years (re: the recent Regents Exam fiasco)?
Lowering standards means more students will pass, but it doesn't mean more will pass that should have passed, and it doesn't mean that education is being improved by allowing to students to be promoted to a higher grade. Once a standard is lowered, the school is then under the gun to show that educational instruction has improved enough for the bar to be subsequently raised. Any school that doesn't demonstrate this is going to have a hell of a time raising their standards again.
The federal standard is for every child to be proficient in reading by 2014. Will we meet that standard? No, because I don't think every child is capable of learning to read proficiently - but I also don't think this goal is too high a goal to set. The purpose of the goal, or standard, in this case is to spur each school and each student to perform to at the top of their capabilities, which is something that perhaps a lower standard wouldn't elicit.