Two reported correlations are discussed in this paper: the first claims a correlation between magnetic polarity and oxidation state of igneous rocks, and the second claims that the Hawaiian hot-spot chain exhibits a significantly dominant normal-polarity bias. On the basis of mechanistic arguments and examination of the sample distribution, the first correlation is shown to be due to a data artifact. Although the second correlation is supported by magnetic modeling of seamounts given in this paper, it too may be due to a data artifact. The most likely alternative explanation is that it originates from a bias imparted by viscous remanent magnetization (VRM) in the Brunhes chron. The role of VRM may be far more important in determining large-scale magnetic anomalies than is generally believed.