‘Realist’ interpretations of personality factors require scales with factorial coherence and independent psychobiological confirmation. Like most (if not all) personality models, the Five-Factor Model traits fail these tests, making them poor candidates for analysing causal relations between personality and either behaviour or life outcomes. More finely-focused scales, such as the facet scales advocated by Mõttus, show more promise for causal analyses, provided they can be shown to be unifactorial, clearly verifiable with psychobiological evidence, and aligned with fundamental psychological theories. Variance decomposition will assist causal analyses using such scales.