Premium
Is it fitting? Comments on the LISREL analysis by Stoner & Arora of variables affecting the psychological health of strikers
Author(s) -
Smith Leigh M.
Publication year - 1989
Publication title -
journal of occupational psychology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.257
H-Index - 114
eISSN - 2044-8325
pISSN - 0305-8107
DOI - 10.1111/j.2044-8325.1989.tb00497.x
Subject(s) - lisrel , psychology , structural equation modeling , econometrics , statistics , social psychology , mathematics
The analysis of the psychological health of strikers presented by Stoner & Arora (1987) presents several problems. The LISREL model that they apparently use is not identified, nor does it fit the publishing data. A reanalysis of the data with an identified model that approximates the authors' model produced parameter estimates which lead to a very different understanding of relationships among the variables. Further analyses showed the data to be empirically under‐identified and therefore unsuitable for LISREL analysis. The processes of model selection, data checking and assessment of fit are discussed.