Regarding the study cited, it all depends on what your exact assumptions are and what kind of population you are dealing with. As we showed, you can get very different results for the various habitancies.
If DRD3ser/ser is a requirement for trait A, then the study's results indicate that trait A is different between the case and control groups. If the habitancy is 100% sanguine, then trait N would be 100% in both the study and control groups, so the difference in trait A would not be due to entanglement (i.e., with trait N actually being the critical difference between the study and control groups). In the context of the NPA model, another thing to consider is whether the difference in trait A is due to association, where the distribution of NPA types is different in the study and control groups. As an example, consider a study where the P trait was causally associated with a disease (say OCD). In a sanguine Polymorphic habitancy, the study group would contain mostly NP and NPA types. Only the control group would contain N and NA types. There would be a modest difference in P trait between the study and control groups (higher P in the study group), but there would also be a difference in A trait (lower A in the study group). Thus, if the researchers mistakenly measured trait A instead of trait P, they could erroneously conclude that lack of trait A was causative in OCD. The difference in A trait here would not be due to entanglement, but due to the preferential partitioning of only NP and NPA+ types (having P trait) into the study group, and all types — but including N and NA types (lacking P trait) — into the control group. That is, the result would be a fluke because of the particular habitancy studied.
« Back to index