Causality: populations, individuals, and assumptions

We side with Lee on the importance and potential use of graphical causal modelling in personality research but raise three issues crucial to its validity. First, causal relations obtained at the inter-individual level should not be confused with intra-individual causal relations. Second, it is difficult to explicate all assumptions about which variables are measured, their causal relations and the possibility of co-occurring events when applying graphical modelling to personality data. Third, multiple testing complicates assessing (in)dependencies. Conclusions and inferences should always be drawn with appropriate attention to the underlying assumptions.