Most Measures are Still Uncommon: The Need for Comparability

  • Version 1.0.0
  • Download 0
  • File Size 231.10 KB
  • File Count 1
  • Create Date August 2, 2018
  • Last Updated August 2, 2018

Most Measures are Still Uncommon: The Need for Comparability

This paper makes explicit the desire and need to link disparate assessments in large-scale assessments. Furthermore, it shows that well established and well-defended studies (like the Anchor Test Study and those associated with the Voluntary National Tests in the US), disagree regarding the viability, validity and practicality of linking disparate assessments. We point out what is now a common procedure in Australia for linking disparate statewide assessments. Furthermore, we present a scenario in India in which we make the argument that such a linking is better than no linking at all. We demonstrate outcomes of one implementation of this linking and share the results both good and bad. We then suggest more research to answer some of the important questions associated with ultimate policy decisions related to this research.Key words: Equating, selection, scaling and psychometrics

Attached Files

FileAction
paper_226dc301c.pdfDownload 
Menu
X