This is Stephen's comment on an ed week blog. Mary Ann Zehr reports on a paper by Robert Liquanti with suggestions for ELL assessment. Mr. Liquanti's paper makes perfect sense if we assume that we LEARN academic language consciously. Stephen suggests that most of academic language is ACQUIRED.
Re: Robert Linquanti, WestEd, Strengthening Assessment
Short URL: http://bit.ly/kt3gQX
Liquanti assumes that academic language can be "defined, taught and measured explicitly" (p. 16). If we accept this assumption, it makes sense to describe the trajectory of skills to be mastered ("mapping out key academic competencies," p. 23) and confirm that students are learning them using formative tests. But this assumption is not supported by research or observation. The arguments and counterevidence have been in the professional literature for decades: The system is very complex: Linguists are still struggling to describe "academic language" and there are no known cases in which people have mastered more than small amounts of academic language through deliberate study.
A better hypothesis is that nearly all of academic language is acquired, or absorbed, the way language in general is acquired: Through understanding message, or comprehensible input. This is confirmed by studies that demonstrate the success of sheltered subject matter teaching, classes in subject matter taught in English, with presentations and readings made comprehensible for second language acquirers. This is also confirmed by studies showing that students acquire an enormous amount of academic language through extensive reading, probably the most powerful tool for reaching advanced levels of English competence. I have argued that extensive pleasure reading is the bridge from lower to higher levels of competence, and brings students to the point where many difficult texts and aural presentations are comprehensible.
If even some of academic language is acquired and not consciously learned, we have to rethink the entire testing situation. If it is true, we should invest in libraries, not more subtle (and expensive) means of testing.
New assessments for the Common Core State Standards Initiative should pay particular attention to how formative tests designed within the assessment systems can improve outcomes for English-language learners, a researcher from WestEd contends in a paper released this week. He says that formative assessments should directly inform teacher instruction and student learning through various practices, tools, and processes.
In laying out his vision for how assessments pegged to the common-core standards could be made valid and reliable for ELLs, Robert Linquanti, a senior research associate for WestEd, a San Francisco-based education research firm, argues that many assessments now in use for accountability under the No Child Left Behind Act are rather rough measurement tools. He says, for example, "that assessment and accountability systems generally treat the [English-learner] category as binary (a student is EL or not), when in fact EL students exhibit language competencies on a continuum that extends from the lowest levels of English proficiency through exit-level performance standards and beyond."
So how to capture that continuum in assessments?
Linquanti explains that the new content assessments are expected to include benchmark assessments that will be implemented at key intervals throughout a school year, formative assessments that can be used to inform teacher practice and student learning, and summative assessments aligned to academic standards.
While the least used in U.S. schools, formative assessments may hold the greatest promise for improving outcomes for English-language learners, Linquanti says. He writes that formative assessments take place within instruction through informal observations and conversations as well as carefully planned methods that give teachers a chance to gather evidence.
Fortunately, Linquanti says, researchers have been making progress in creating formative assessments relevant for ELLs. He cites a project implemented by the World-Class Instructional Design and Assessment consortium. It's called FLARE, which stands for the Formative Language Assessment Records for English-Language Learners. He also mentions a WestEd professional development program called Quality Teaching for English Learners that includes formative assessments>. (I featured this program in EdWeek last spring.)
Linquanti says future assessments should integrate this kind of professional development model that enables teachers to evaluate if ELLs are meeting language and content objectives.
Linquanti also makes the point that before students are given an assessment, they must have a chance to learn the material. He emphasizes that ELLs have to get an opportunity to acquire the discipline-specific English language they need in the first place. That means that all teachers must have expertise to help students develop the specialized vocabulary, sentence-level structures, and discourse patterns needed for the particular content areas they teach, he says.
Linquanti's suggestions for how ELLs can best be included in the next generation of content assessments were published in a report, "The Road Ahead for State Assessments," released yesterday by the Policy Analysis for California Education and the Rennie Center for Education Research & Policy. My colleague Catherine Gewertz blogged about it over at Curriculum Matters.
Linquanti hasn't written his paper in a vacuum. He mentions a number of heavyweights in the ELL field who gave him feedback on his vision before it was released. Last month, I quoted some of those experts in an article for EdWeek about some of the issues state consortia face in designing new assessments to include ELLs.
Along with the paper on assessments for ELLs, the report includes a paper on computer adaptive assessments and one about assessing students in science. "None of these topics has received the attention that it deserves in the current debate on assessment policy," the paper says.