Testimony Before the Senate Committee on Education on the
End-of-Course Testing Provisions of SB2
Angela Valenzuela, Ph.D.
Education Committee Chair
Texas League of United Latin American Citizens
June 23, 2005
Thank you for this opportunity to speak to you today. I am here representing Texas LULAC. I also testify before you as an expert in the area of student assessment. I've edited a recently published volume, a book on assessment and accountability in Texas centered around a very explicit concern that I and many others across the state and nation have. And this concern is about the use of high-stakes tests-of which end-of-course exams are about. If you read the research in my edited volume titled Leaving Children Behind conducted by scholars like Linda McNeil, Richard Valencia, Kris Sloan, Elaine Hampton and others, it becomes clear that many of the problems that we are having in our schools here in Texas are generated by the testing system itself-specifically, the high-stakes testing component.
Before getting into specifics, I want to address the underlying assumptions that typically remain unspoken in order to question the logic of this system. Two primary assumptions are that this is a good business model and secondly, that it embodies a civil rights approach. These assumptions hold with respect to both the current TAKS system of testing and the end-of-course one that is being proposed.
As a business model and using the market as a metaphor, it is illogical to make either students or teachers responsible for the quality of their product when they do not control the resources or flow of finances to which the outcomes are tied. With respect to children and their parents, when in business do we make the consumer responsible for the quality of the product? We are moving forward with the implausible premise that a level playing field in the quality of personnel and instruction exists or will exist across all courses offered in every school and in every classroom in our state in order to justify a standardized examination process.
Regarding Civil Rights, the case of de-segregation is instructive. This history reminds us that when the statutory engines of change commanded restraint, they targeted not the victims of Jim Crow, but rather the perpetrators of discriminatory practices. Just as it would have been unreasonable-indeed, nonsensical-for either the lowest-level workers or business clientele to have assumed the primary burden of change to integrate lunch counters, hotels, and other public establishments, so too is it unreasonable for both children and their teachers to bear the primary burden of change under the banner of accountability.
Now to specifics, the problem isn't with standardized tests; when used appropriately, these are legitimate forms of assessment and no author disputes the need for accountability. The issue is with the way that the tests are used. Specifically, they are used to accomplish various goals simultaneously, compromising its validity. First, they are used to measure achievement (albeit in a narrow way and also in a way that is insensitive to the needs of language minority and disabled children-the latter by the Commissioner Nealey's own admission); second, as primary determinants of children's progression within the system (which even the test makers themselves say it should never do); and third, to monitor or control the behavior of the adults-mainly teachers but also principals-in the system.
Here is the rub. The problem here is that the testing system doubles as both a testing and monitoring system-two functions in one. One cannot parcel out from a single indicator the extent to which that number was achieved through excessive coaching or other forms of system gaming. And so when we tie course credit to end-of-course examinations, perverse incentives to narrow the curriculum by teaching to the test will logically occur. All of this, of course, corrupts the assessment so that the score you get never reflects real achievement.
As Education Committee Chair for Texas LULAC, my studied recommendation is that the Senate Committee on Education remove the end-of-course provisions from your bill. Not only has there been no public discussion of this significant direction that you want to take us in, I see no scholarly justification for additional high-stakes exams. In fact, I see the very opposite: There is an ample research base, coupled with myriad newspaper reports from, and on, our own state that clearly indicate the wrong-headedness of this approach.
I could mention, for instance, the recent cheating scandal discovered by the Dallas Morning News, affecting approximately 400 schools across our state. Although that actual number as been adjusted to a lower figure, to not infer dysfunction is to willfully ignore this evidentiary base.
More positively, I'll point to the most comprehensive study to date conducted by Stanford professor Linda Darling-Hammond, titled, “Multiple Measures Approaches to High School Graduation.” She conducts an analysis of 27 states that differ from single-test approaches like Texas in that they consider a variety of student work, including student academic records, research papers, portfolios, essays, oral exams, and capstone projects. In general, she finds that multiple measures approaches have both raised achievement and lowered dropout rates. In Texas, there is evidence of increasing test scores, but every other indicator that we care about (ACT, SAT, high school completion) is running in the opposite direction.
Other good news is that under NCLB this is actually one area-student assessment-over which we can exercise our own independent judgment as a state. We cannot do anything about the high stakes at the school and district level-but, as the Darling Hammond study indicates-we do have latitude with respect to high-stakes decisions involving individual children. When so much is at stake-particularly for the children themselves, I urge this committee to proceed with utmost caution and to reconsider the addition of more high-stakes testing. Thank you very much.