You’re a good math student learning English as a second language. You take a standardized test that poses proportion problems involving coins. Because you never learned that the term “one quarter” can refer to “a 25-cent piece” as well as “one fourth,” you end up scoring poorly.  

With English language learners soon to comprise more than 50 percent of the nation’s K-12 school population, that scenario – presented in October at the 2014 Teachers College Columbia Roundtable on Second Language Studies by Carolyn Wiley, a senior research scientist with the Educational Testing Service (ETS) – reflects a fundamental truth that schools and districts must confront on a daily basis: school assessments need to begin to think about the role of language.

“Language use is situated in context,” says conference organizer James Purpura, Associate Professor of Language and Education. “Second language learning is never just about learning language forms, but always involves other factors such as other kinds of knowledge. That means that when test takers are using a language other than their first language, assessment in any discipline needs to account not only for understandings of the disciplinary context but also differences in communicative English language ability.”

The goal of the third Teacher College, Columbia University Roundtable in Second Language Studies (TCCRISLS), was to “bring together scholars, researchers, teachers and students interested in the interfaces between assessment and second language learning,” Purpura said. The event drew more than 150 scholars in applied linguistics, psychology, educational measurement, and other fields from the United States, Europe and Asia, and delegations from assessment research hubs such as ETS and the University of Cambridge (both of which were among the co-sponsors). Many presenters were not strictly based in language education, but rather worked on issues around assessment in subjects like reading, science and mathematics.

There was widespread agreement that current formal assessment does not always put the learner at its center. Standardized large-scale tests, given their use in selection and in their current form, do not adjust to individual learner needs as they perform on tests. Formative assessments – in-class exercises or discussions designed to narrow learning gaps, for instance – may capture whether students have picked up specific information, but, if poorly implemented, may not show whether, and how well, that information has been integrated into the more complex process of learning.

A growing field known as Learning Oriented Assessment (LOA) seeks to address assessment design, implementation, and validity inquiries. However, the promise of LOA is tempered by the fact that assessing language learning is particularly complicated given the fact that language is not only to object of assessment, but also the medium of assessment, and so many factors contribute to performance.

“Most research in mainstream assessment is in ‘content’ areas like math and science,” said Carolyn Turner, Associate Professor of Second Language Education at McGill University. “But L2 [second language] learning is even less linear” and less about amassing a certain body of knowledge”, than about helping learners proceduralize the language knowledge they have in order to communicate in a wide range of contexts.

Constant Leung, Professor of Educational Linguistics at King’s College London, addressed what makes instructor feedback in assessment work or fail. “Many teachers find student take-up of their feedback unpredictable and often disappointing,” Leung said, before delving into what he called the “disciplinary hinterland that lurks in questions and feedback.”

Leung presented examples from a master’s degree class in TESOL in London, categorizing responses to feedback as “rejection,” “happy let-it-pass,” “critical acceptance” and “fulsome reception.” A learning-oriented approach, he said, should anticipate this mix of reactions and design assessment methods that help to bring all students on board in a “learning space where feedback can lead to adaptation.”

Other papers focused on applications to assessment design or results from pilot studies. Mikyung Wolf and Alexis Lopez of ETS, for instance, discussed the use of incorporating scaffolding into assessment of English language learners’ proficiency, based on a study of a diverse sample of 140 students in grades K-5. Scaffolding, understood as providing supports to help students complete particular tasks, can take many forms, from questions that help build the logic behind the task to role-playing or interactive exercises. Digital technology can help build these tools in increasingly appealing ways that adapt to each student, for instance appearing only when needed. Such design also “offers a promising means of engaging students in learning during assessment,” they noted.

In another paper, James Pellegrino, distinguished professor at the University of Illinois-Chicago, discussed how the Common Core State Standards challenge designers of reading assessments to capture ability to “read and learn and acquire information from multiple sources” – not just books but multiple media, online sources, and so on, involving skills such as evaluating hypotheses or verifying data. Teachers themselves are not always fully prepared for this shift, he noted – a fact that assessment designers must accommodate as well.

Taking stock, Purpura said we are still far from having a set of best practices for learning-oriented assessment in general, let alone in second-language settings. A new TC/ETS Forum on Teaching, Learning and Assessment of English Language Learners, which will meet three times a year starting in 2015, is aimed at addressing that gap. Purpura believes the effort will benefit both from new digital technologies for making scenario-based assessments and a growing interest from large-scale assessments to the learning realities of the classroom.

“The field of language assessment now recognizes that there is a paradigm shift,” Purpura said. “Previously we looked at achievement and placement tests, and ignored what transpired in the classroom and how it related to learning. Now we are focusing on the nature of assessment in the classroom – assessments that are embedded in instruction, instruction and assistance embedded in assessments, student responses while learning, feedback and assistance, and obviously more formal achievement tests. Now they all come together for the purpose of learning.” – Siddhartha Mitter

Published Thursday, Feb. 26, 2015