R. M. Baron and D. A. Kenny (1986; see record 1987-13085-001) provided clarion conceptual and methodological guidelines for testing mediational models with cross-sectional data. Graduating from cross-sectional to longitudinal designs enables researchers to make more rigorous inferences about the causal relations implied by such models. In this transition, misconceptions and erroneous assumptions are the norm. First, we describe some of the questions that arise (and misconceptions that sometimes emerge) in longitudinal tests of mediational models. We also provide a collection of tips for structural equation modeling (SEM) of mediational processes. Finally, we suggest a series of 5 steps when using SEM to test mediational processes in longitudinal designs: testing the measurement model, testing for added components, testing for omitted paths, testing the stationarity assumption, and estimating the mediational effects.

Most empirical tests of mediation utilize cross-sectional data despite the fact that mediation consists of causal processes that unfold over time. The authors considered the possibility that longitudinal mediation might occur under either of two different models of change: (a) an autoregressive model or (b) a random effects model. For both models, the authors demonstrated that cross-sectional approaches to mediation typically generate substantially biased estimates of longitudinal parameters even under the ideal conditions when mediation is complete. In longitudinal models where variable M completely mediates the effect of X on Y, cross-sectional estimates of the direct effect of X on Y, the indirect effect of X on Y through M, and the proportion of the total effect mediated by M are often highly misleading.

The multitrait-multrniethod (MTMM) design is often used in test validation research to disentangle problems due to shared method variance. However, MTMM research requires extensive data collection that may be prohibitive in clinical settinp. Furthermore, interpretation of MTMM data can be ambiguous and misleading-In the current article, maximum-likelihood confirmatory factor analysis (CFA) is presented as a means for less ambiguous interpretation of complete and incomplete MTMM designs. Confirmatory Factor analysis is applied to four data sets that represent four designs: monotrait-monomethod, multitrait-monomethod, monotrait-multimethod, and multitrait-multimethod. In all four cases, CFA results provided more rigorous support of the original authors' positions and provided valuable supplemental findings as well. Additionally, CFA was applied to an artificial data set. Intuitive interpretations of the artificial data were compared with the results of the CFA. The results showed that the intuitive approach can lead to highly spurious conclusions regarding convergent and discriminant validity.Campbell and Fiske's (1959) multitrait-multirnethod (MTMM) approach provides, perhaps, the best available test of validation research. Ideally, the clinical researcher hopes to make statements about the convergent, discriminant, and, ultimately, the construct validity of the selected measures. In practice, however, two stumbling blocks exist. The first pertains to the feasibility of MTMM data collection, especially in clinical settings. The expense, the time, and the intrusion on both clients and clinical staff may result in more modest data collection (e.g., measuring only one trait with multiple methods or measuring multiple traits with only one method). The second impediment concerns data analysis. Even if a full MTMM correlation matrix is obtained, how should it be analyzed? A variety of data analytic approaches have been advocated.Of course, the first was the visual comparison of zero-order correlations described in Campbell and Fiske's (1959) criteria.Other approaches included use of the analysis of variance (ANOVA) paradigm (originally proposed by Guilford, 1954; subsequently embellished by Stanley, 1961), a nonparametric proximity function (Hubert & Baker, 1978), a partial correlation method (e.g., Schriesheim, 1981), and exploratory factor analysis (e.g., Golding & Sneidman, 1974;Tucker, 1966). More recently, however, maximum-likelihood confirmatory factor analysis (CFA) has been recommended for use with MTMM matrices (Schmitt & Stults, 1986), And yet, in spite of these ad-1 thank the following people for their valuable comments on preliminary drafts of this article:

No abstract

The latent trait-state-error model (TSE) and the latent state-trait model with autoregression (LST-AR) represent creative structural equation methods for examining the longitudinal structure of psychological constructs. Application of these models has been somewhat limited by empirical or conceptual problems. In the present study, Monte Carlo analysis revealed that TSE models tend to generate improper solutions when N is too small, when waves are too few, and when occasion factor stability is either too large or too small. Mathematical analysis of the LST-AR model revealed its limitation to constructs that become more highly auto-correlated over time. The trait-state-occasion model has fewer empirical problems than does the TSE model and is more broadly applicable than is the LST-AR model.