Multimethod factor analysis in the evaluation of convergent and discriminant validity.

Hierarchically Nested Covariance Structure Models for Multitrait-Multimethod Data

“…For example, the statistical tests derived by Hubert and Baker (1978) are not sensitive to the particular pattern of correlations among traits, though Hubert and (1959) criteria, and have proposed alternative approaches to evaluating multitrait-multimethod data. Some developed one-step (Jackson, 1969(Jackson, , 1971; but see Conger, 1971) or two-step (Golding & Seidman, 1974;Jackson, 1975; but see Golding, 1977, Jackson, 1977 component analytic approaches to identifying trait and method factors. Others have applied structural equation modeling procedures to multitrait-multimethod data.…”

Comparison of Three Methods of Assessing College Environments1

“…Because of a number of limitations in evalu ating multitrait-multimethod matrices, Jackson (1969) recommends multimethod factor analysis as a method of examining convergent and dis criminant validity. He suggests eliminating method variance from multitrait-multimethod matrices by "orthogonalizing the diagonal monomethod mat rices prior to a principal components analysis and rotation of axes [1969, p.…”

“…A new technique that Jackson (1969) refers to as multimethod factor analysis was explored in this study in order to examine common variance across methods and to investigate the convergent and discriminant validity of those institutional variables purportedly measured by more than one method. Campbell and Fiske (1959) argue that methods or teste should converge in their assess ment of the same trait; moreover, if the method of assessment is to be considered independent of a given trait or variable, it must also show dis criminant validity.…”

Validity

“…Although not technically an analytic procedure in itself, several powerful analytic methods have been appended to it (Jackson &: Paunonen, 1980;Schmitt, Coyle, &: Saari, 1977). These include nonparametric (Hubert &: Baker, 1978), analysis-of-variance (Kavenaugh, MacKinney, &: Wolins, 1971;Stanley, 1961), factor analytic or principle component (Golding, 1977;Golding &: seidman, 1974;Jackson, 1969Jackson, , 1975, confirmatory factor analytic (Kalleberg &: Kluegel, 1975;Kenny, 1976;werts, .1'6reskog, &: Linn, 1973), and path-analytic (Schmitt, 1978; werts, J'oreskog, &: Linn, 1972; werts &: Linn, 1970)' models and techniques. Moreover, since method effects may be multiplicative rather than additive, some direct-product modeIe have also been formulated -:-74 - (Browne, 1984;campbell & O'Connell, 1967).…”