OECD Programme for International Student Assessment (PISA)
Publication Date
9-2008
Subjects
Student surveys, Research methodology, Questionnaires, Databases, Test construction
Abstract
This paper examines trends between PISA 2003 and PISA 2006 for the reading and mathematics scales for the OECD countries using both the standard one parameter (1PL) model that has been used in published OECD reports, and a generalised partial credit model, as implemented in ACER ConQuest, which estimates optimal values for each possible score on an item.
Recommended Citation
Macaskill, G. (2008). Alternative scaling models and dependencies in PISA. Australian Council for Educational Research (ACER). https://research.acer.edu.au/pisa/12
Place of Publication
Melbourne, Victoria
Publisher
Australian Council for Educational Research (ACER)
Comments
TAG(0809)6a Prepared by Greg Macaskill for TAG Meeting, Sydney Australia 29 September – 2 October 2008