The hypothesis-based investigation of patterns of relatedness by means of confirmatory factor models: The treatment levels of the Exchange Test as exampl

This paper presents a method for formalizing hypotheses on the outcomes of performance according to different treatment levels that can be investigated by means of confirmatory factor models. This method includes several steps starting from the general research hypotheses and the characteristics of the stimulated processes and ends up with a pattern of relatedness that can be integrated into a confirmatory model for an investigation. It is a fixed-links model adapted to the pattern of relatedness with the factor loadings constrained according to this pattern. This method is demonstrated in a measure of working memory capacity that includes several treatment levels. Three alternative hypotheses and the corresponding patterns of relatedness are considered for an investigation. The data for the investigation originate from an internet study. The results demonstrate the usefulness of this systematic approach to the investigation of assumed patterns. It turns out that in the internet data one pattern is clearly superior to the other patterns