- Nov 9, 2022
- 755
- 1,156
Deepest, and most abject and tearful apologies....but you would need to construct and analyze a contingency table to see if these variables are really and truly independent after taking enough data. It is possible, for example, that the probability of a person being abusive, given that they are dominant is higher than the probability of being abusive given that they are not dominant. It might be interesting to also bring submissiveness into the mix. In order for the variables to be completely independent we would need, where D=dominant and A=abusive:Being Dominant, and being abusive are not mutually exclusive things. It's possible to be both abusive and Dominant. It's possible to be Dominant and not abusive. It's possible to be neither dominant nor abusive. It's possible to be abusive but not Dominant. They are independent unrelated variables.
P(D)=P(D|A) and P(A)=P(A|D), that is, the probability of a person being dominant = the probability of a person being dominant, given that they are abusive. And the probability of a person being abusive = the probability of a person being abusive, given that they are dominant.