Experts on assessments for students with disabilities recently weighed in on how the U.S. Department of Education (DOE) should spend $350 million in Race to the Top Fund monies. The funds have been designated to support the development of new assessments with a common set of standards.
The DOE has held a series of input meetings across the country on the Race to the Top Fund program not only to inform the federal government of what it should include in its request for proposals, but also to guide states in their efforts to design new assessments and submit applications for the funds.
Input meetings have been held on general assessments, high school assessments, assessments for English Language Learners and on technology and innovation in assessment.
At the meeting on assessments for students with disabilities held in Atlanta in November, several experts in assessment provided their thoughts, comments and recommendations. Below are brief summaries or exerpts of the experts who spoke:
Martha Thurlow, Director, National Center on Education Outcomes
In the early 1990s, most states included 10% or fewer of their students with disabilities in their state assessments and NAEP had less than 50% of students with disabilities participating. Accommodation policies were either non-existent or limiting.
Today nearly all states have more than 95% of their students with disabilities participating in their assessments and it’s made a significant difference in their education. But, because of a history of restricted access to the general education curriculum, students with disabilities aren’t performing well.
The assessments, in general, are showing us the reality of being a student with disability in US schools today. Just as there is with general education students, there is a wide range in performance for students with disabilities.
I think it’s irresponsible to assume that because a student has a disability that student is a low performer who cannot learn. When states first began to include their students with disabilities in assessment, several were startled to learn that their students with disabilities increased the overall performance levels in the state.
One of the challenges is creating a system that doesn’t punish them for having profiles that may differ from most kids. You’re likely to find kids who don’t have some of the basic skills like decoding, but do have higher level skills, say, in math. They may not be able to do the adding and subtracting very well, but they are great at problem solving.
I’ve identified 3 basic requirements that the RFP should address:
1) Develop assessments with all students in mind. Too often in the past we’ve been retrofitting our assessments for students with disabilities, and we shouldn’t allow that to happen at this time of opportunity. There should be a recognition that students with disabilities are general education students first. We should expect at least 80% of all students with disabilities to meet the same achievement standards as students without disabilities.
Whether we call it universally designed assessments or accessible assessments, it’s critical that during the development process we think of all students and clearly define what the assessment is intended to measure. Assessments should be designed in a way that allows all students to show their knowledge and skills.
2) Define allowable accommodations. Defining content standards and clarifying what we’re trying to measure is so important as we think about accommodations.
For example, the ability to decode is a separate skill from the ability to draw inferences from text. Students may be poor decoders but fine comprehenders. We can’t tell that from our assessments, and we’re actually penalizing students because we’re putting those two together.
To the extent that content targets are not clear, or that they intertwine things such as decoding and understanding of a text, we should think about how to separate those. Too often, students with disabilities are penalized because their disability is in an area that is the focus of the assessment.
The National Accessible Reading Assessment Projects has focused on the development of some principles and guidelines to help states think through what they need to have in accessible reading assessments. Ensuring that we have common standards that we address accessibility concerns does not mean lowering the standards.
3) All students need to be assessed, no exceptions. We know from past history and research that when any group of students is excluded from the assessment system, there’s always the next group of students on the precipice of being excluded. The group of excluded students becomes larger and larger and when you are excluded from assessments you also are likely to be excluded from quality instruction.
Some of my bottom lines are that we should do nothing that would reopen doors to lowered expectations and outcomes for students with disabilities; we should do nothing that could result in a separate curriculum for some students with disabilities; and we should do nothing that would track students into a separate achievement expectation early in their school careers.
Lizanne DeStefano, Professor, University of Illinois College of Education
It is always amazing to me how quickly people shift to the assessment as the driver of the system, and the assessment is what’s important. But when we really think about the theory behind standards-based reform, it’s the standards that are driving the system, not the assessment.
So I would hope that that would come across very clearly, and that a phase of development of these projects be examining the standards upon which this system is going to be based, and making sure that these standards are sound and also represent the idea of inclusivity.
Many times when I look at state systems I see that at the “meets or exceeds” category, you’re talking about what students can do, but the categories below “meets” often focus on the deficits of those students rather than what those students can actually do. And as a teacher, I want to know what the strengths of my students are, what they are able to do, as well as areas of challenge.
We need adequate precision throughout the performance continuum. If we’re really going to design assessments that are inclusive and represent the full range of students, then we have to pay attention to how precise are they at all points in the performance continuum.
But if we are really going to provide useful information across the full spectrum, then we need to be assured that our test is measuring with good precision at all levels. In some assessments that are designed to high standards, precision in the lower quartile is poor. They have more items that measure at the upper end, fewer items at the lower end, and students in the lower end of the performance continuum don’t have the opportunity to get very many items right on the assessment.
One of the things that I’ve been working on for the last few years is thinking about a modular assessment where every student takes a common block, so there’s a set of items that every student takes, and then there’s an adaptive block that a student takes that is chosen based on their prior performance. We’re really not talking about making the test easier, or dumbing down the test. What we’re talking about is an adaptive system that allows us more precision at the extreme ends of the curriculum.
I would shift the language away from minimum competency or basic competency and think about a learning progression or a learning continuum and think about that as anframework for performance descriptors.
Jacqui Kearns, Director, National Alternate Assessment Center,
University of Kentucky
In a survey that we just finished in 7 states, we found that 75% of the 1% (students with significant cognitive disabilities) who are participating in state alternate assessments are reading sight words and using a calculator to do basic math operations. Now, the sad thing about that is they’re doing that in elementary school, middle school and high school. There is no curriculum progression. They are also symbolic language users that either use oral speech or they use a symbol-based augmentative communication system to communicate.
One of the things I’d like to emphasize is authentic demonstrations of skills and knowledge for al students. Authentic contextualized demonstrations are essential for students with significant cognitive disabilities. In order for them to generalize information there has to be a context. Include structured processes that ensure that all have t he opportunity to show what they know. And by structured processes, I’m thinking there may be some scaffolding required.
If you understand standards-based assessments, and you understand standards-based curriculum, any kid with a disability will be successful in their classroom. I’m really, really excited about the possibility of giving teachers feedback and information on students in real time, during the course of the school year.
One of the things that we’ve been able to do in the last couple of years with the assessment data in Kentucky is use the information we learned from the student results, as well as the characteristics of the learners to inform professional development and leverage dollars under other grant programs.
I really like Lizanne’s (DeStefano) comment about informing teacher training and licensure and involving higher education from the get go as a part of the process. A very critical part of the process is building their capacity along the way for what you’re going to do and why you’re going to do it.
So what I would encourage you to do is to be very thoughtful about who the population is. Don’t just guess that you have 1% or 2%. In your design process really think very carefully about the student population, the content, and the theory of learning that goes with that content.
We found the assessment triangle of Pellegrino, Chudowsky and Glaser’s, in their book, Knowing What Students Know,to be very useful in the work we did with early alternate assessment. It helped us to think about who the student population is, the cognition vertex, the observation vertex, and the interpretation vertex, and those three pieces being balanced.
You know, one of the things I think happens in large-scale assessment is we expect assessments to tell us everything. Well, they can’t. We have to be specific about what we need from them and how they should be used, and what the resulting information can and cannot do. I think it’s critically important that we are clear about that because the validity of those assessments hangs on the clarity of that.
Brian Gong, Executive Director, National Center for the Improvement in Educational Assessment (NCIEA)
The first thing is that we should build on the gains made in assessment for students with disabilities over the past 20 years. The RFP has to address more advanced assessments, but it should take care to do no harm.
The requirement that students be tested on grade level is a way to force — try to force–access to the general curriculum. For students with disabilities that’s a much bigger issue than it is for many other students. And so I would advocate for a clearer distinction between assessment and accessibility than for regular assessment for regular students.
People sometimes talk about monitoring and driving the system. It’s very hard to have assessments that do both, that provide good information about monitoring and then also drive for aspirational things when we talk about the monitoring part, but just a reminder that the driving part is really important.
Everyone has been talking about growth models and progress and learning progressions like these are new things, but they aren’t new things for educators of students with disabilities. These are very different traditions than the standards-based accountability that we’re obviously talking about.
What are the common standards for students with disabilities, including students with the most severe cognitive disabilities? The standards must include more detailed content and information about expertise than current performance standards do. If we’re trying to assess for the full range of knowledge and skills, we have an under-specification in the current content standards.
Assessment design should not be divorced from a theory of action of how the information will be used.
For the full transcript of this hearing, go to:http://www.ed.gov/programs/racetothetop-assessment/atlanta-transcript-2.pdf