Equating tests across grades in large-scale assessments: A study of two alternate methods in conjunction with methodological innovations

Sonia Sappl

Research output: ThesisMaster's Thesis

253 Downloads (Pure)

Abstract

The validity of vertical equating relies on the common items, present in the different Year-level tests,
showing fit to the model and exhibiting no differential item functioning. In this thesis, two empirical
data sets are re-analysed to investigate different approaches to conducting vertical equating for
investigating item functioning using the Rasch model. The major objectives are: to study the
efficiency of a simultaneous analysis of both year groups compared to separate analyses of each
year group; and to investigate a novel approach to equating where only the common items are
analysed initially, followed by a simultaneous analysis of all items.
Original languageEnglish
QualificationMasters
Awarding Institution
  • The University of Western Australia
Supervisors/Advisors
  • Andrich, David, Supervisor
  • Houghton, Stephen, Supervisor
Thesis sponsors
Award date11 Jul 2023
DOIs
Publication statusUnpublished - 2023

Cite this