ABSTRACT
In this section I want to consider some data relating gender and physical activity to obesity, shown in Table 1. I am going to approach this as though gender and inactivity are causes of obesity. It is worth noting, however, that even if this is wrong, there is little or no harm in performing the analysis as if it were correct. Causal models are also associational models, so that those who do not want to make causal interpretations can fail to do so. The reverse is not the case, since most associational models are not causal, so that the use of a purely associational model can prevent a causal interpretation.
