The term "statistical significance" is often misunderstood or abused to imply a large effect size. A recent example is in the work of J. P. Rushton (1988, 1990) on differences between Negroids and Caucasoids. Rushton used brain size and cranial size as indicators of intelligence, using Pearson "r"s ranging from 0.03 to 0.35. These statistical facts indicate that the relationship of brain size to intelligence is too weak and inconsistent to use brain size as a measure of intelligence. Based on similarly modest relationships in some studies of race to brain size and crime, he classified Negroids as genetically inferior to Caucasoids. Due to methodological errors and a poor review of literature, he also rated Negroids as inferior on other variables such as mental health. In summary, Rushton's writings about human races are statistically unfounded. One table is presented, derived from Rushton. (Author/SLD)