Scholarlink

An important methodological challenge for survey assessments is how to manage the requirements to evolve and innovate with respect to assessment content, test delivery, and analysis methods while maintaining the core purpose of providing valid trend results. In NAEP, one way this challenge has been addressed over the years is through the conducting of special linking studies that use a random groups design (Kolen, 2007, Data collection designs and linking procedures. In N. J. Dorans, M. Pommerich, & P. W. Holland (Eds.), Linking and aligning scores and scales (pp. 31–55). Springer). In this chapter, we describe an operationally feasible approximation of NAEP standard errors that incorporates an estimate of the linking error. We present the method developed and used in the reporting of the 2017 NAEP results. We provide an illustration of the method using the 2015 and 2017 NAEP results, including a comparison to an exact formula-based method appropriate for group means. Lastly, we describe ongoing theoretical and practical work that builds on and improves the methods used in 2017.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Approximate Standard Errors for NAEP Results That Incorporate Linking Error Under the Random Group Design

  • 章节

  • 发表日期：

    2025年

查看全文
  • John Mazzeo,
  • Bingchen Liu,
  • John Donoghue,
  • Xueli Xu

摘要

An important methodological challenge for survey assessments is how to manage the requirements to evolve and innovate with respect to assessment content, test delivery, and analysis methods while maintaining the core purpose of providing valid trend results. In NAEP, one way this challenge has been addressed over the years is through the conducting of special linking studies that use a random groups design (Kolen, 2007, Data collection designs and linking procedures. In N. J. Dorans, M. Pommerich, & P. W. Holland (Eds.), Linking and aligning scores and scales (pp. 31–55). Springer). In this chapter, we describe an operationally feasible approximation of NAEP standard errors that incorporates an estimate of the linking error. We present the method developed and used in the reporting of the 2017 NAEP results. We provide an illustration of the method using the 2015 and 2017 NAEP results, including a comparison to an exact formula-based method appropriate for group means. Lastly, we describe ongoing theoretical and practical work that builds on and improves the methods used in 2017.