Towards a rationale for screening test in reading in first grade – construct, format and cut off

Towards a rationale for screening test in reading in first grade – construct, format and cut off

[featured_image]
  • Version
  • Download 541
  • File Size 266.48 KB
  • File Count 1
  • Create Date October 30, 2019
  • Last Updated October 30, 2019

Towards a rationale for screening test in reading in first grade – construct, format and cut off

As the development of adequate reading skill is considered fundamental for further academic and life success, teachers and policymakers need high quality screening test to early on identify students in need of extra follow up. Increasingly over the past decade there has been a demand of more thorough documentation of the quality of national assessments (e.g. Evers et al., 2013; Arnesen et al., 2018), often precipitating a focus on psychometric standards. Different from diagnostic tests and tests developed for research purposes on an individual level, group administered tests screening children for reading difficulties tend to base design and cut off on available concurrent data and often focus on more isolated aspects of reading. In this study we argue for a better concurrence of prevailing reading theory and measures of reading. Further, in the litterature there seem to be no consensus for how cut-offs are developed and what is more, there are few qualitative benchmarks to guide cut offs for identifying children in need of support in developing adequate reading skill.

In this study we follow the lines of Kane (2017) who states that there is no such thing as a correct cut-score, however reasonable cut scores are applied as the appropriate criterion of quality. The approach of Kane (2017) follows the so called Goldilocks principle which entails that “the standard need to be high enough to achieve the goals of the program, but not so high as to cause serious side effects.” (Kane, 2017, p. 28). When putting this much value on reasonableness, Kane moves an argument based approach to validity (Chronbach, 1988, Kane, 2013) at the forefront. This, however, does not exclude other aspects of validity, as validity is considered to be unitary, i.e no aspect of validity is more important than the other.

Attached Files

FileAction
20190185_3199_Towards a rationale for screening test in reading in first grade_.pdfDownload