An example demonstrates that Joreskog's (1969) suggested sufficient condi
tions for identifying unknown parameters in a confirmatory factor analytic model
with correlated factors are not sufficrent. Sufficient conditions for identifying
parameters in the class of models used by Burt et al. (1978) to estimate population
structures of well-being from various indicators of satisfaction are given. The
six hypotheses stated in Burt et al. are reconsidered and similarly supported when
their final covariance model is restricted so as to meet the sufficiency conditions
given here.
Get full access to this article
View all access options for this article.
References
1.
Bielby, W.T. and R.M. Hauser (1977) "Structural equation models." Annual Review of Sociology, Vol 3. Palo AltoAnnual Reviews.
2.
Burt, R.S. , J.A. Wiley, M.J. Minor, and J.R. Murray (1978) "Structure of well-being: form, content and stability over time." Soc. Methods & Research6: 365-407.
3.
Dunn, J.E. (1973) "A note on a sufficiency condition for uniqueness of a restricted factor matrix." Psychometrika38: 141-143.
4.
Howe, W.G. (1955) "Some contributions to factor analysis " Report ONRL-1919 Oak Ridge. Oak Ridge National Laboratory
5.
Joreskog, K.G. (1969) "A general approach to confirmatory factor analysis ." Psychometrika34: 183-202.
6.
Long, J.S. ( 1976) "Estimation and hypothesis testing in linear models containing measurement error a review of Joreskog's model for the analysis of covariance structures " Soc Methods & Research5: 157-206
7.
Sorbom, D. (1978) "An alternative to the methodology for analysis of covariance." Psychometrika43: 381-396.
8.
Werts, C.E. , K.G. Joreskog and R.L. Linn (1973) "Identification and estimation in path analysis with unmeasured variables." Amer J of Sociology78: 1469-1484.
