FairTest's Monty Neill's Remarks at the FairTest-NEA Conference

These remarks were delivered by Monty Neill, Ed.D., Interim Executive Director of FairTest, as part of a panel discussion following a keynote presentation by Dr. James Pellegrino.

 

Thank you. I am very happy to be with you today, and I am looking forward to a very interesting, challenging and productive conference.


I'd like to make two main points:

    First, the need to shift from an accountability model to a shared responsibility approach. 

    Second, the necessity to build systems that rely primarily on school and classroom assessing.


Accountability:


Jim Pellegrino has explained well the limits of the tests, and why we need to change the assessments we use. 


Accountability as now conceived is highly centralized, top down, and acts as a 'gotcha' tool. As a result it causes narrowing of curriculum and instruction to the tests and inhibits the kinds of work necessary to create high quality schools.


We can redefine accountability or we can use better terms, such as shared responsibility. In either case, we need a different approach that does the following:


    1. Uses multiple sources of evidence of student learning across all important areas.


    2. Includes opportunity to learn data from both schools and communities (such things as health care availability).


    3. Focuses on school improvement efforts and successes. In various documents from the Forum on Educational Accountability that you have in your packet we discuss some of that, especially professional learning and parental involvement and support. [See Redefining Accountability and Assessment and Accountability for Improving Schools and Learning, both on the web at www.edaccountability.org.]


    4. Uses the multiple sources of evidence across inputs, improvement efforts, and results to evaluate school status, efforts, and improvement or lack of improvement.


    5. Provides to schools – and communities – that which has turned out to be lacking and needed.


    6. Builds systems to foster improvement. Such systems must be based fundamentally on schools as communities of learners engaging in shared practice. 


    7. Engages in targeted interventions when evaluation of data shows there are problems and schools, despite help, are not improving.


This means responsibility is shared among governments, educators, parents and communities. It requires continuing dialog and mutual respect. Only in a context of shared responsibility, rather than enforced top-down accountability, can the energy of educators truly be released.


Local assessment:


A healthy evaluation of student learning will draw on many forms of assessments. I agree with Jim: we have to find ways to develop new tools, good tasks and projects, etc., and professional learning will be essential. I want to emphasize some other aspects. Some of this is in the expert panel on assessment report in your packet [see links above], which Jim and Alba Ortiz and I worked on. This report is the basis of the starting principles for our discussion. I would emphasize the following:


    - First, teachers are the primary assessors. They need access to many tools and must know how to use them well, but in any event a good deal of assessment in a rich, high-quality classroom necessarily means assessing on the fly, adapting, deciding to use assessment B instead of A, etc., replying to the emerging needs of each and every student. It is a core teaching skill. In addition, teachers should know how to create good assessments – not because any one teacher can create all the good assessments she will need, but because that knowledge is necessary for understanding assessment and because good assessments will need to bubble up from teachers. In sum, we must respect teachers as assessors but ensure they can become good at this work.


    - Two, the kinds of information that come from the ongoing flow of classroom work provide essential data for being able to fairly and helpfully evaluate individuals and schools. Thus, assessment information should draw on evidence from three sources: ongoing classroom work; the flexible use of approved questions, tasks and projects; and larger-scale assessments. By approved questions, I mean, for example, a bank of tasks that teachers can use in instruction or for more formal assessing done as part of classwork (rather than on the state education department's schedule). Technology is beginning to provide many valuable tools and procedures to make this workable.


    - Three, there needs to be flexibility in using assessments. The requirement that all students answer the same questions or perform the same tasks at the same time should be minimally employed with little weight relative to the evidence provided by flexible use of approved items and tasks and information from the ongoing flow of classroom work. But this flexibility must be embedded in schools in which educators share and develop their practice.


    - Four, states must develop systems that rely on the three sources of evidence and a variety of types of assessments to provide public information and support thoughtful evaluations. Note that I am emphasizing evaluation by humans, not judgments triggered automatically by a set of test scores.


    - Five, there are several reasons a statewide system must include local evidence. If, for example, performance tasks are a key part of the evidence of learning, as they should be, we cannot realistically have many tasks shipped off for central scoring. That means we must have local scoring, as is done in many nations. (There are ways to ensure quality and consistency.)

    In addition, the desire to ascertain progress in multiple subjects runs up against the danger of far too much testing (e.g., annual testing in multiple subjects). Relying on locally controlled, largely classroom-based information, can solve that contradiction.

    To avoid a rigid system that is bound not to work for many, we need local flexibility in deciding which assessments to use when. Assessments also must be incorporated into instruction so classes don't have to constantly stop to do external tests, such as the currently popular "benchmark" or "periodic" tests.

    As I noted above, teachers are the primary assessors. By expecting local assessments to include classroom evidence from the regular flow of student work and by building in the supports to enable strong teacher-based assessing, we commit to systems that really are about employing high-quality professional educators. That seems to be the secret in Finland and other more educationally successful nations.


Finally, ensuring that local assessments, varied as they will be, are of adequate quality and that different does not mean low expectations for some, will take time to work out. The evidence I have seen tells me that acceptably uniform and accurate systems can be built. If we must trade some looseness in data as a price for a system that does not turn schools into relatively low-level test prep, that should be an easy choice to make. Further, if the emphasis is on using evidence for improvement, not for gotcha and sanctions, a modest increase in looseness creates no real problems.


In sum, we need high-quality, mainly performance assessments, we need local flexibility within reasonable bounds, and we need a shared responsibility approach to school improvement.


Thank you.