Premium
Easier Said Than Done: Rejoinder on Sijtsma and on Green and Yang
Author(s) -
Davenport Ernest C.,
Davison Mark L.,
Liou PeyYan,
Love Quintin U.
Publication year - 2016
Publication title -
educational measurement: issues and practice
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.158
H-Index - 52
eISSN - 1745-3992
pISSN - 0731-1745
DOI - 10.1111/emip.12106
Subject(s) - cronbach's alpha , reliability (semiconductor) , internal consistency , consistency (knowledge bases) , psychology , odds , epistemology , statistics , mathematics , computer science , psychometrics , philosophy , clinical psychology , artificial intelligence , power (physics) , physics , logistic regression , quantum mechanics
The main points of Sijtsma and Green and Yang in Educational Measurement: Issues and Practice (34, 4) are that reliability, internal consistency, and unidimensionality are distinct and that Cronbach's alpha may be problematic. Neither of these assertions are at odds with Davenport, Davison, Liou, and Love in the same issue. However, many authors in the testing community mention these terms not only together, but sometimes as if they are synonymous. Moreover, Cronbach's coefficient alpha is very popular as an index of reliability. Thus, articles discussing alpha are not only appropriate, but necessary. Our concerns are the same as formed the genesis of prior (2009) articles by these same authors, Sijtsma and Green and Yang. This rejoinder also makes comments about item parcels when tests are multidimensional and about factor analytic approaches to assessing reliability.