Scientific realism has become exceedingly popular as a frame of reference, especially within research drawing upon structural equation models. This concept allows for modification of a causal model derived from theoretical considerations based on preliminary empirical findings. This is significant as such an approach increases the risk of ending up with a model that misleadingly signals a high degree of validity and reliability. It is, therefore, imperative to interrogate whether the indices which reflect the fit of the model are consistent with alternative samples. This paper addresses the aforementioned problem. Additionally, the goal here is to examine to what extent the bootstrapping technique contributes to the evaluation of the stability of global fit indices.