|
|
David
D Curtis
Australian Council for Educational Research
curtis@acer.edu.au
Peter
Boman
School of Education, James Cook University
peter.boman@jcu.edu.au
 Read complete article
|
Abstract
By using the Rasch model, much detailed diagnostic information is
available to developers of survey and assessment instruments and to
the researchers who use them. We outline an approach to the analysis
of data obtained from the administration of survey instruments that
can enable researchers to recognise and diagnose difficulties with
those instruments and then to suggest remedial actions that can
improve the measurement properties of the scales included in
questionnaires. We illustrate the approach using examples drawn from
recent research and demonstrate how the approach can be used to
generate figures that make the results of Rasch analyses accessible to
non-specialists.
Rasch, partial credit model, reliability, threshold
analysis, differential item function
|
Back
Install
Acrobat Reader