The Testing Begins... and a call for accountability.
Anyway, you heard me right, it's January, and we're taking high stakes tests. Some seventh graders will repeat the grade or spend their summer in remedial classes as a result of what they did today. This is because it takes the city/state (it's a state test, but I think that scoring it is the responsibility of the city) over five months to score it, compile results, and notify students who failed about their options for promotion.
Thinking this through... school started in September. That means we have had approximately 4 1/2 months to prepare the children for a test that will have a very real impact on their lives. Apparently, the state considers that enough time to teach the students a year's worth of reading, writing, and listening skills. I realize that I am glossing over the post-test months at the end of the 04-05 school year, but that's because we can all agree that this year's teacher can only be responsible for what happened this year, right? So anyway, 4 1/2 months. But the powers-that-be require 5 1/2 months to get the results ready. Does this seem wrong to you in any way? It seems a little messed up to me that the government/testing companies gets longer to score a test (something which can be done largely by machine once the essays are marked) than the teachers and students get to learn the material, yet it is schools, teachers, and students who bear the brunt of the call for accountability.
I'm all for accountability. I'm not reflexively against standardized testing, in limited quantities and used in conjunction with other measures of progress. But I think it's time that we turned the tables on the test-makers & test-mandators. The movement started with mr. e:
With all that is riding on these tests, you would think that the tests would be well written and accurate, or at least proofread. Unfortunately, that is not always the case. So I'm promoting a new movement for all of those know-it-alls outside of education to jump on -- accountability for test writers.He continues with a list of concerns about this year's 4th grade test. Building on his ideas, let me propose a few principles of accountability for high-stakes testing:
- Tests must be error-free. It is completely unacceptable for even one question to contain errors. I'm not talking about subtle issues of phrasing, here, either. I have heard rumors (can anyone confirm?) that on the fourth grade ELA test, in one story, a main character's name changed mid-way through the story! And today, we were instructed to discard the testing instructions for the sixth grade test because they were incorrect and had to be replaced with a new set of instructions. Furthermore, a few minutes after we started today's test, my principal opened her email to find an URGENT email about a problem with the seventh grade test. I don't know what it was, as I was a hall monitor and could not hang around talking, but I watched my AP rush down the hall muttering about how this would never be accepted in a more affluent area... (Unaccountable Talk has the details: The letters assigned to the answer choices did not match up with the letters assigned to the bubbles on the answer document. Don't forget that 7th grade is a new automatic-holdover year, and confusion over five questions could lead to a kid - hundreds of kids! - spending another year in middle school. This is not okay). Every subgroup of testing document must be error-free.
- Test results must be available to students, parents, teachers, and administrators within two months, maximum. We live in a computer age. If you budget two weeks for marking essays, a week for feeding answer documents into scanners, and a week for analyzing results, I don't see any reason why it should take more than 1 month to turn around these tests. Assuming that there are 1000 reasons why that timeframe is impossible, I think 2 months should be the absolute limit. Test results must be available to all within two months. Period. Not just ELA and Math, either, I include the Science and Social Studies tests in this mandate. No test left behind.
- No one should score standardized tests unless they are paid extra to do so. Yes, they are paying teachers to score the ELA tests over February break, and that's fine. But as far as I know, we are all scoring Math tests for two days this spring while the kids get a long vacation. That's not what I signed up for when I went into teaching. If I need some extra cash, I'll sign up for scoring sessions, but I'm sure as heck not going to do it for free. High quality tests cost money: to produce, to field test, to print and copy-edit and deliver and administer, to score. It is time that the government faced the real costs of all this testing. If you want to give a test every year to every child, fine, but you're going to have to pay for it. And that includes paying people a reasonable hourly wage to score it. Don't think you can pass that off on your teachers; we have enough to do already, and this wasn't our idea.
- The testing schedule should be based on the most appropriate timeframe for judging student learning, not on the needs of the companies and governments administering it. No high stakes testing in January, or February, March, or April. Well, okay, late April to early May might be reasonable given an allowance of two months to process the results and notify families about summer school. Again, the government's scoring & reporting problem should not become my students' problem. It is unfair to judge the work of a teacher or school - let alone a student - on 4 1/2 months of work. Tests must be given as late as possible in the school year in order to reflect the knowledge and skills gained that year.
- Tests must be the highest-quality assessments available, no matter what. That means performance exams for Science. It means carefully field-tested questions in all areas. It means scoring by people with at least a few hours (paid) training. It means checks for scoring consistency from one region of the state to the next (there are rumors that we actually score harder here in the city than they do in the suburbs). It means rigorously checking questions for vague phrasing, class bias, and other forms of unfairness. We are judging human beings with these instruments, sorting people, deciding who stays with their peers and who gets held back, who gets admitted to gifted programs and selective schools and who doesn't, who gets evaluated, remediated, labeled, promoted... how could we use anything less than the most sophisticated instrument available? Oh, right: that costs a lot. Not my problem. If the tests are important, find the money. Only highly-qualified tests need apply.
- Money for testing shall not be taken out of existing education budgets. This may be the most important item on the list, given that I keep repeating that a good testing program, done right, isn't going to come cheaply. It also must not come at the expense of actual education. Again, if testing is something important to the American public, we ought to find the money to pay for it, above and beyond that which we already allocate to schools.
Taxpayers, how can you expect less?
Problems with tests ought to be highly publicized, like the lists of failing schools. Don't provide any details; just list the name of the testing companies, the names of the responsible departments within the city and state government, and list 'em under the headline: Failing. And then - do schools and teachers and students get to transfer out, to find alternative ways to assess?
Spread the word!