Monday, March 10, 2008

Pop Quiz on Testing, Senate Bill 1 and the definition of "all kids"

There seems to be broad agreement that Kentucky school students should be tested. State schools invest millions of dollars and thousands of hours per year assessing learning. But there is much less agreement about why students should be tested.

This is a problem. But, it is not really a “testing problem.”

This is a political problem. How does a community of responsible adults care for all of its children?

Let’s start with a pop quiz.

Question: Public school students should be tested…

a.) often; so that a teacher might catch learning problems early, adjust instruction and report to parents how well their children are learning specific objectives

b.) annually; in order to inform parents whether or not their children are “college material”

c.) annually; in order to hold schools accountable for the performance of all of their students

d.) all of the above

e.) a & c only

Have you got an answer in mind? (Weigh in on KSN&C's Pop Quiz, top right)

Broad agreement on the answer to this question would go a long way toward resolving the issues surrounding Senate Bill 1 – which proposes replacing the CATS assessment with a national multiple-choice exam.

Senate Bill 1 is sponsored by Senate President David Williams and Majority Floor Leader Dan Kelly. It is of questionable constitutionality but that’s not a major obstacle for an experienced - constitutional - interpreter like Williams.

Kelly explained the motivation behind Senate Bill 1 to KET’s Bill Goodman recently on Kentucky Tonight saying he wanted to “make sure that all students have the opportunity to find out ...whether or not they are college material. Some, we think, will find out they are, when maybe their family background would say that they aren't. And some, like my children, who thought they were, might find out they aren't." I suppose Kelly would choose answer “b.”

On the surface this answer seems OK. As a parent, I certainly wanted an idea of how my children were performing relative to others and normative data should be a part of CATS. As a principal, I wanted that data too. But it’s not just about "my child." It’s about "all" of Kentucky’s children. If SB1 is tied in any way to Kentucky’s historic tendency toward acceptance of inequitable education, it needs to be stopped.

The present circumstance has been brewing since KERA required the monumental task of creating a performance-based assessment system that ranked and judged schools, rather than children. From the beginning, one of the hardest things for parents to understand (second only to the multi-aged primary program) was why students would sit for days of testing and when it was over, parents did not learn much about how their children were performing. They were used to testing that focused on their child.

As scientific testing became increasingly popular over the last century the main instrument for group assessment has been the nationally-normed multiple-choice exam, like those suggested in SB1. Such tests were designed to tell parents how their children performed on a set of objectives (that may or may not have been taught in their classroom!?) and compared that performance to the average scores of other children of the same age.

But the tests were also designed to discriminate among students; to separate the “college material” from those who would work in the fields and factories. Viewed narrowly, it gave parents a useful metric for determining how their own child was doing relative to other kids. But viewed broadly, it confirmed deeply held beliefs about which students were worth educating. While it might serve a majority of top students fairly well, Senate Bill 1 would return Kentucky to the days when it was expected that many students would not achieve at high standards. That’s why many educational leaders have tended toward answer “c.”

Teachers have always given assessments in their classrooms to find out, “Who got it” and “Who didn’t.” This is testing in its purest form; diagnosis and prescription for student learning. It begins with the curriculum – what you want students to know or be able to do – and ends when the teacher is satisfied the student knows or can do it. No political agenda - just an effort to improve a child’s learning. I suspect most teachers would prefer answer “a.”

Testing today is like the CTBS on steroids. In addition to teacher-made tests, school districts use other assessments to get snapshots of how students are performing in advance of the state’s CATS assessment – the ultimate yardstick for measuring results in Kentucky schools. The already complex CATS assessment was recently made worse when NCLB required unreasonable targets for the schools.

Teachers should recognize that any efficient system needs inertia and keeping a reasonable amount of “pressure” on school personnel to perform is a good thing for children; as long as it is fair, well-focused on the needs of all of its students, and leaves teachers better prepared to be successful.

We need a comprehensive, accurate and stable accounting of student progress over time - including an improved accounting of dropouts. There is a place for normative assessment within CATS but the state would be foolish to surrender Kentucky's curriculum by outsourcing it to an out-of-state vendor.

But, as it is, CATS needs work. Anytime schools have to derive one of their two accountability scores from a concordance table - to see what they would have scored on a test their students didn't take - look out!

There is much to talk about.

Commissioner Jon Draud has called for a Task Force to look into it. Then, for reasons that are unclear to me, the Courier-Journal called him a “wimp.” Perhaps C-J mistook Draud for "Railroad" Commissioner.

For citizens who want schools to be accountable, Senate Bill 1 is not the answer. It would waste precious time going in the wrong direction? The state would have to figure out a new way to describe proficiency and school goals - based on a “yardstick” that always leaves half of the children below average; even when they have made great gains. Surely encouraging growth must be central to any efficient system.

Normative data is for individuals and should remain a minor player in state-wide accountability – not its central feature. There remain practical limitations to any test’s ability to describe what a human knows or can do and claims of “great advances” in multiple-choice testing technology since 1990 look more like purposeful fabrication to me. But some folks swear by it. I think schools should get out of the business of telling students that they will never amount to anything. Instead we should do everything possible to double the number of college graduates in Kentucky.

For that reason, the best answer for all of Kentucky’s children is “e.”

This is an age of information and technology. The old “industrial model” of the school-as-social filter is morally bankrupt and economically suicidal and must be avoided at all costs. Senate Bill 1 is not the solution to Kentucky's problems.

No comments: