A St. Petersburg Times Editorial assumes that Florida's Governor Bush is "gloating" over the recent decision to keep test items out of the hands of parents:
An appeals court has now determined that Florida parents don't have a right to see where their students are going wrong on standardized tests, but Gov. Jeb Bush shouldn't be so eager to gloat. His win comes at the expense of students who are being held back without really knowing why.
Why assume Governor Bush is "gloating" about this? Withholding test items is nothing to gloat about; it's merely standard testing practices with companies and/or school districts that do not have the money to create new forms each year.
The governor portrayed his opposition to disclosing FCAT test materials as consistent with "the Department of Education's 20-year policy on test confidentiality," but that's a little disingenuous. Until four years ago, DOE never used a standardized test to grade and punish schools. Until last spring, it never used a state test of such complexity to decide whether high school seniors could graduate. Until this fall, it never used a state test to decide whether third-graders should be promoted or retained.
The switch to using a test for these methods does not necessarily compel the test developers to release test forms. High stakes do not require disclosure. Some companies do disclose test items, but only because they have the staffing and the funding to create new test forms each year that have been equated and checked for bias and validity issues. I've commented on the lack of understanding of the financial issues surrounding testing before.
And a test "of such complexity" for seniors? Please. Here are 10th-grade Reading sample items. The reading passages are appropriately difficult, but the questions are often very easy. For the first passage, simply knowing the meanings of the words "surly" and "maxim" gets a student two right answers out of eight multiple-choice items. The specifications for these passages indicate that "how-to" articles and advertisements are appropriate "forms of informational text" on the exam; this suggests that some reading passages might be very, very easy.
The technical report (p. 16) notes that the p-values for the 2000 Operational FCAT are distributed as might be expected for a test of basic skills. For example, on the 10th-grade reading test, there are a few hard items, but 75% of the items were answered correctly by at least 59% of the examinees.
And did I mention that Florida's students get six chances in high school to pass the 10th-grade FCAT?
The issue of using a test to promote third-graders is thornier. The little kids get fewer passes at the test, and one could legitimately argue that students that young are not disciplined enough to deal with a high-stakes standardized test, nor has anyone come up with a solution for what to do with the ones who flunk this test repeatedly.
Because Florida has so dramatically increased the stakes associated with one test, it owes students and teachers a better understanding of how they are performing. It also owes them better assurance that testing error didn't lead to grave consequences in their lives.
I agree. But this "better assurance" doesn't mean driving the costs of tests up astronomically by releasing items to parents. It means implementing quality control mechanisms to ensure that the test is error-free before it is administered, to ensure that the scoring and score reporting processes are bug-free, and to ensure that scores are being interpreted correctly (i.e., with a standard error of measurement band that reflects the reliability of the test).
Posted by kswygert at December 2, 2003 02:48 PM