In this week’s post, I’ll share some information related to assessing English language learners that I gathered while attending the Council of Chief State School Officers National Conference on Student Assessment last week. You can take a look at the CCSSO NCSA conference program as well as several presentations here.
This conference was very timely, as the full operational administration of the CCSS content assessments led by Partnership for Assessment of Readiness for College and Careers (PARCC) and Smarter Balanced Assessment Consortium (SBAC) is taking place in the 2014-15 academic year. With the CCSS assessments quickly coming down the pike, there was quite a lot of talk about them at the conference. It was also refreshing to note that the unique considerations of ELLs as well as Students with Disabilities were a frequent topic of presentations.
My big takeaway from the conference is that a lot of research is being done on assessment, including on assessment of ELLs, but I’m concerned about how the findings are reaching teachers. To help spread the news to teachers, I’ll do my best to share some salient information on assessment for ELLs in this post. This post will include information on formative assessment of ELLs, Next Generation assessments for unique populations, ELLs with disabilities, and what’s happening with English language development assessments.
Formative Assessment and ELLs
In a very interesting presentation by WIDA’s Gary Cook, Dr. Cook first provided some background information on factors that impact the assessment of ELLs. For example, he said that teachers are providing interventions with the wrong information about ELLs, and he spoke about the importance of educators knowing more about their students to be able to deliver quality instruction and assess them equitably. He shared also some demographic information which impacts assessment, such as the fact that 57% of ELLs are born in the United States. In addition, while 70% of ELLs in the US are in urban areas, in states that are part of the WIDA consortium, 70% of school districts have fewer than 100 ELLs.
Dr. Cook stresses that when we assess ELLs, we need to assess “construct relevant language” or language that is critical to the content. Although it depends on where ELLs start, Dr. Cook’s research finds that it takes 5-7 years for ELLs to learn English construct relevant language, which is why assessments need to be designed to measure the appropriate content. In essence, we need to understand what the language is that we want these kids to know and design assessments around it. To that end, every item on SBAC has a linguistic complexity level assigned to motivate that consortium to check the level of construct relevant language presented.
When asked his opinion about his level of confidence in the assessment industry when it comes to ELLs, Cook said he was encouraged that SBAC and PARCC recognize the issue. However, he feels the assessment vendors may have a longer way to go to get that message.
Next Generation Assessments for Unique Populations
Another fascinating presentation by staff from Measured Progress and ETS captivated my interest by actually demonstrating test items that allow for unique populations such as ELLs and Students with Disabilities to better access them. (I highly encourage you to look at the helpful visuals used in this presentation!) Presenters described how the essence of Universal Design for Learning (UDL), which frames the design of the PARCC and SBAC assessments, is flexibility and the inclusion of alternatives. They demonstrated that a digital test item isn’t just one fixed piece of content – like it would be on a paper and pencil test – but rather a complex system of possibilities that allow for students with different strengths and needs to better demonstrate understanding of concepts assessed.
For example, the presenters demonstrated how one test item can be assessed using supports such as Braille, a video clip of American Sign Language, and a translation of the item. Despite the advances being made in assessment, unless students have prior instructional experience using these types of digital supports, they can be confusing.
One feature of digital items that resonated with me is that the stigma of having an accommodation can be greatly reduced with computerized testing. When ELLs use accommodations with paper and pencil tests, such as a bilingual glossary, it’s obvious to other students that ELLs are using this support. With technology based testing, however, all the accommodations are built in to the computer interface and specialized for individual students. It won’t be apparent who is using accommodations and who is not. Therefore, more students eligible for accommodations are apt to actually use them. On a related note, PARCC recently approved its testing policies for ELLs including information about accommodations.
ELLs with Disabilities & Assessment
I also learned about the Improving the Validity of Assessment Results for English Language Learners with Disabilities (IVARED) project, which is funded by a federal grant and awarded to a consortium of five state departments of education (Arizona, Maine, Michigan, Minnesota, and Washington) with the National Center on Educational Outcomes. The project aims to learn more about how ELLs with disabilities (also called dually identified ELLs) learn best and how teachers can most effectively support their learning. IVARED is: (1) identifying each state’s population of dually identified ELLs and relating that information to the students’ assessment performance, (2) describing inclusions of ELLs with disabilities in state assessment participation and accommodation policies, and (3) developing principles and guidelines for assessing these students.
ASSETS and ELPA21
As you probably are aware, two consortia are creating English language proficiency/development (ELP/D) standards and assessments that are correlated to the language found in the CCSS. A presentation at CCSSO centered on ELP/D assessments and was titled Challenges of Transitioning English language Proficiency/Developmetn Standards and the New On-Line Assessment System to Reflect the Language Behind the Common Core. The lineup included presenters representing WIDA (Tim Boals, Gary Cook, Margo Gottlieb, Carsten Wilmes), ELPA21 (Kenji Hakuta), Pearson (Edynn Sato), the Center for Applied Linguistics (Dorry Kenyon) along with the Wisconsin Department of Education (Audrey Lesondak) and the California Department of Education (Lily Roberts).
The presenters shared that WIDA’s Assessment Services Supporting ELs through Technology Systems (ASSETS) will contain some notable differences over their current ACCESS for ELLs assessment of ELP/D. Since ASSETS will be a technology-based assessment system, it will contextualize the source of the dialogue or text that students work with. In writing, there will be a variety of stimulus that presents the academic content to contextualize it for students. For the speaking portion, students will have a virtual test administrator and virtual student who models the tasks students are to take part in. The Center for Applied Linguistics has been conducting cognitive labs of more than 115 students to study the mental processes students use when completing ASSETS tasks and have received encouraging results so far.
As for ELPA21, Dr. Hakuta focused more on the English language development standards piece that is not funded by the grant funds but that will be assessed on ELPA21. ELPA21’s ELD standards build on California’s ELD standards but take these standards another step forward. ELPA21’s standards do so because the CA standards are aligned to English language arts only and the ELPD Framework wasn’t officially released when CA developed their standards. The WestEd team is working on developing the standards with Dr. Lynn Shafer Willner leading the charge.
You’ve heard enough from me! What have you been seeing in your school, district, or state in terms of how ELLs will be assessed on the CCSS and on English language development? How does formative assessment fit in?