Citation

Desai S, Braitman KA. J. Fam. Violence 2005; 20(2): 101-107.

Affiliation

Department of Psychology, Southern Illinois University, Carbondale, Illinois; Department of Psychology, Boise State University, Boise, Idaho; Texas Health and Human Services Commission, 909 W. 45th Street, Building 2, Mailcode 2010, Austin, Texas, 78751

Copyright

(Copyright © 2005, Holtzbrinck Springer Nature Publishing Group)

DOI

10.1007/s10896-005-3173-8

PMID

unavailable

Abstract

Little research has examined the properties of survey instruments when the full instrument is not administered (scale carving). The purpose of the present study was to test scale carving effects on survey instruments. Participants completed carved and whole versions of two instruments measuring violent behaviors and attitudes at two sessions. Findings indicated that administering select subscales rather than the entire instrument affects psychometric properties of certain instruments. For instance, mean scores of certain instruments differed depending upon whether the instruments were administered in their entirety or in carved versions. Reliability coefficients did not differ between carved and whole versions, and correlation coefficients were moderate. Therefore, psychometric properties of carved instruments should be assessed prior to their use in research.