A Confirmatory Approach to SEM

Testing a SEM model is said to take a confirmatory approach. Put another way, a conceptual model is determined a priori and then data is collected to test how well the model fits the data, thus trying to “confirm” the researcher’s hypotheses about how constructs influence one another. Joreskog (1993) outlines how SEM testing usually follows one of the following approaches in confirmatory model testing.

1. Strictly Confirmatory Approach—the researcher has a single model that is accepted or rejected based on its correspondence to the data. This could be considered an “all or nothing” approach where either all of the model is accepted, or it is fully rejected.This is a very narrow approach and is rarely

Figure 1.4  Strictly Confirmatory Approach

2. Alternative Models Approach—more than one theoretical model is suggested, and you test to see which one has a superior fit (Figure 1.5). Most of the time, researchers are hard pressed to find two alternative models supported by the literature.

Figure 1.5 Alternative Models Approach

3. Model Development Approach—a model is initially tested using SEM and any area of the model that is non-significant or ill-fitting is changed based on the data. Hence, an alternative model is suggested based on the data analysis. These suggestions are often based on modification indices or analysis properties that point out where stronger relationships exist. This is by far the most common method. There is a drawback to this approach because these “developed” or modified models can be hard to The suggested changes to a model may just be an artifact of a specific data set. Thus, you are often capitalizing on chance if your data set is determining your model conceptualization. Once a “developed” model is found to be accept- able, it is a good idea to achieve another data set to verify the structure of the revised model.

Source: Thakkar, J.J. (2020). “Procedural Steps in Structural Equation Modelling”. In: Structural Equation Modelling. Studies in Systems, Decision and Control, vol 285. Springer, Singapore.