> I wonder how the researchers control for the effects of external factors Since no one else mentioned it: Religiosity is my guess for the key factor. In parti

> I wonder how the researchers control for the effects of external factors Since no one else mentioned it: Religiosity is my guess for the key factor. In particular "married couple who had not lived together before marriage" would be strongly correlated with faith. Religiosity -> Marriage and Religiosity -> Stability.