ABSTRACT

This article presents a critical review of the Texas English Language Proficiency Assessment System (TELPAS), a large scale standardized English language proficiency (ELP) assessment developed by the Texas Education Agency (TEA) and administered since 2004. TELPAS is used as an annual summative assessment for all English Learners (ELs) in grades K-12 and is administrated annually to approximately one million ELs in Texas. It is a high-stakes assessment impacting students, campuses and districts in multiple ways including EL reclassification decisions, campus and district accountability, language program evaluation, and instructional/assessment accommodation decisions. TELPAS has not been critically reviewed previously and little research has investigated its impact on the large EL population in Texas or other stakeholders. TEA has recently made many structural changes to TELPAS making this an ideal time to review the assessment. The current review showed TELPAS to be a sound instrument but there is room for improvement. Recommendations include increased frequency of released tests to allow for external reviews of the assessment, steps to take to allow for a more rigorous inter-rater reliability study, an in-depth reliability study of the new machine scoring of listening and speaking, and a robust consequential validity study.

摘要

本文介绍了对德克萨斯州英语水平评估系统（TELPAS）的严格审查，该系统是由德克萨斯教育局（TEA）开发并于2004年开始实施的大规模标准化英语水平（ELP）评估。TELPAS被用作年度总结对K-12年级的所有英语学习者（EL）进行评估，并且每年在德克萨斯州管理着大约100万英语学习者。这是一项高风险评估，以多种方式影响学生，校园和地区，包括EL重新分类决定，校园和地区责任制，语言课程评估以及教学/评估适应性决定。以前尚未对TELPAS进行严格的审查，很少有研究调查其对德克萨斯州或其他利益相关者中大量EL人群的影响。TEA最近对TELPAS进行了许多结构性更改，使之成为审查评估的理想时机。当前的审查表明，TELPAS是一种声音乐器，但仍有改进的空间。建议包括增加发布测试的频率以允许对评估进行外部审查，采取步骤以允许进行更严格的评分者间可靠性研究，对新机器的听力和口语评分进行深入的可靠性研究以及鲁棒性结果效度研究。