. . "The random assignment of children rules out the chance that factors other than thimerosal, such as education or poverty, caused the results." . . .