Mediation Analysis via Observation Oriented Modeling

Psychologists routinely treat psychological attributes as quantitative in nature, and the majority of statistical tools traditionally employed by psychologists also rest upon the assumption of continuous quantitative structure. There is no evidence, however, that attributes such as intelligence, depression, personality traits, etc. are truly quantitative, thus rendering most statistical techniques inappropriate for psychological data. Observation Oriented Modeling (OOM) is an alternative data analysis technique that does not require the assumption of continuous quantity, and it is an approach that eschews aggregate statistics while returning individual observations to the forefront of the analysis. Two published mediation models are re-analyzed in the current paper using OOM and the results compared to those obtained from traditional, variable-based analyses. It is shown that OOM yields results that are entirely transparent and readily interpretable. Unlike traditional analyses, the results also provide a clear understanding of what is happening at the level of the persons in the two studies. OOM is therefore a novel way of conceptualizing and analyzing data that may lead to more accurate and complete explanations of psychological phenomena.