Development of ROAR-Español: Created, Validated, and Normed for Spanish-Speaking Multilingual Learners

• In Event: Leveraging Research-Practice Partnerships to Develop Open-Access and Scalable Reading Assessments: Rapid Online Assessment of Reading

Abstract

Objective
This talk will provide an overview of the development of ROAR-Español, a suite of Spanish foundational reading skills designed as an extension of the ROAR platform. We will also present results from validation studies, made possible through sustained research-practice partnerships.

Motivation and Theoretical Background
Spanish-speaking multilingual children represent a large population of students in the U.S. However, their foundational reading skills are often solely assessed in English, which risks underestimating their full literacy abilities (Thordardottir et al., 2006). It is thus recommended that multilingual children are assessed in more than one language whenever possible to support more accurate and equitable decisions for instruction (ASHA, 2004). To attend to this need, we developed ROAR-Español, a Spanish version of ROAR, to tap into Spanish-speaking multilingual learners’ reading skills. The development was guided by collaborative partnership with schools to ensure the assessment was culturally and instructionally relevant.

Development of ROAR-Español
ROAR-Español consists of four assessments Letra, Fonema, Palabra, and Frase that were developed from the core assessments (Letter, Phoneme, Word, and Sentence, respectively) of the English version of ROAR. Throughout the development process, we prioritized cultural sensitivity and ensured that linguistic characteristics of Spanish were accurately portrayed and assessed. Collaboration with experts from diverse Spanish-speaking regions provided critical input to item development and assessment instructions.

Validation and Contextualization of ROAR-Español
The validation of ROAR-Español was made possible through our strong partnerships with schools and research partners. Spanish-speaking multilingual learners took both the ROAR-Español assessments and traditional reading assessments (e.g., Woodcock-Muñoz Language Survey) that are considered the “gold standard.” Our research partners also provided locally significant assessments that allowed a regionally appropriate validation as well. There were two phases of the validation study where we first worked with a predominantly monolingual Spanish speaking population from Colombia. Next, we recently extended the validation study to multilingual learners within the U.S. In both validation studies, we found a strong relationship between ROAR-Español Frase and Woodcock-Muñoz (Colombia concurrent validity: N=639, r=0.87, p<.001; U.S. concurrent validity: N=121, r=0.60, p<.001)
Effectively serving multilingual learners does not end at test development—appropriately contextualizing assessment scores is imperative for practical use. Based on feedback from our research partners, providing benchmark scores for multilingual learners on a monolingual population scale is most impactful in practice. These normative scores, which provide a baseline for where a multilingual learner is in relation to their monolingual peer, are valuable for practitioners to distinguish between the effects of foundational reading skills and language backgrounds, enabling instruction to be appropriately tailored.

Significance
The inherently complex nature of multilingualism underscores the importance of mindfully creating assessments that attend to the unique linguistic profiles of multilingual learners. The development, validation, and norming of ROAR-Español was possible through a research practice partnership model to create culturally responsive and empirically grounded tools. Future work will integrate English and Spanish scores to support holistic, strength-based interpretations of ROAR-Español.

Authors

• Mia Fuentes-Jimenez, Stanford University

• Kelly Rose Wentzlof, Stanford University

• Wanjing Ma, Stanford University

• Julian M. Siebert, University of California - San Francisco

• Ana Saavedra, Stanford University

• Carrie Townley Flores, Stanford University

• Adam Richie-Halford, Stanford University

• Jason D Yeatman, Stanford University