今日の洋書

• McNeill, WH, "Plagues and Peoples"

Moreover, it is worth considering the psychological implications of a disease that killed only Indians and left Spaniards unharmed. Such partiality could only be explained supernaturally, and there could be no doubt about which side of the struggle enjoyed divine favor. The religions, priesthoods, and way of life built around the old Indian gods could not survive such a demonstration of the superior power of the God the Spaniards worshiped. Little wonder, then, that the Indians accepted Christianity and submitted to Spanish control so meekly. God had shown Himself on their side, and each new outbreak of infectious disease imported from Europe (and soon from Africa as well) renewed the lesson. (pp.20)

• Hosmer & Lemeshow, "Applied Logistic Regression (Wiley Series in Probability and Statistics - Applied Probability and Statistics Section)"
  • Confounder と effect modifier (interaction)

Note that the coefficient for gender changed from 1.356 to 4.239. This is not surprising since the inclusion of an interaction term, especially when it involves a continuous variable, usually produces fairly marked changes in the estimated coefficients of dichotomous variables involved in the interaction. Thus, when an interaction term is present in the model we cannnot assess confounding via the change in a coefficient.