Tuesday, December 8, 2009

Thoughts on Race to the Top: Collaboration and local control

By now, grant applications for the federal Race to the Top (RTT) program are in preparation across the country. There are a lot of state department of education employees who are likely to have a hardworking holiday season as applications for the first wave of funding are due in January. With the 4.3 billion dollars that has been allocated, states will have significant resources that can be brought to bear to make the kind of sweeping and dramatic updates and changes to school systems that are called for as part of this federal initiative.

The guidelines presented to the states to prepare their RTT applications contained two clear themes. On the one hand, state education initiatives are supposed to preserve the “flexibility and autonomy” of LEAs. There is clear recognition of the need for districts to be supported in their efforts to make decisions about curriculum assessments and other issues that are in the best interests of their staff and students. In addition to the call for local control there is also a clear mandate for collaboration. States are encouraged to adopt common standards and collaborate to produce common assessments. One of the questions that state governments face in their preparation of their proposals is how best to balance these two, at times seemingly contradictory, objectives.

One of the ways that collaboration could be facilitated, while at the same time preserving the decision making power of the districts is for the state to make available to districts an item bank in which all of the items are on the same scale. These items could be used on both district interim assessments and the state test. What would this mean for the sake of districts you ask? Such an item bank would afford the opportunity to make sure that the assessments composed of these items, either entirely or in part, can be placed on the same scale. This means that the scores are directly comparable. The 500 on the math test given in the middle of the year could be compared directly to the 550 on the state test at the end of the year. Put another way, in this case, the statement could be made that the ability level required to achieve 550 on the benchmark test is higher than the ability level required to achieve 500 on the state test. Without tests that are on a common scale, such comparisons are not possible. The 550 might represent higher ability than the 500 and then again it might not. Having a common measuring stick could go a long way towards facilitating collaborative work.

A common item bank could also greatly assist smaller districts in their efforts to implement valid and reliable interim assessments for the purpose of informing instruction. The utility of assessment results is greatly aided to the extent that they reflect the kids that actually attend the district schools and the instructional priorities of the district. Research has consistently shown that the items behave differently when students change or when instruction changes. Ongoing analysis of test behavior is critical to making sure results are reliable and valid for the kids with whom it will be used. Such analysis is difficult with districts that have only small numbers of kids. Having a common item bank from which to draw could make it much easier to do the necessary analyses to back up an assessment initiative with a small district.

Achieving these beneficial results does not require that both the state and district tests are comprised entirely of items from this bank. The only requirement is that they both contain a sample of items from the bank. This would leave the district free to include local items reflecting content which may not be of interest to other districts in the state. Easy communication and collaboration need not be sacrificed in order to continue to achieve the flexibility and autonomy that allows districts to make sure that instructional improvement systems meet their priorities.

Anyhow, I had best sign off at this point. This post is already rather lengthy. We would, as always, be interested in hearing the feedback of others about these ideas or about other topics.

No comments: