Just published online in Environmental Health Perspectives is a study attempting to link failed in vitro fertilization with PCBs. But the researchers ignored the classic multiple comparisons problem. Here, 57 congeners were tested and only one produced significantly elevated dose-dependent odds of failed implantation — less than you’d expect by chance at a 5% error level. Moreover, the association between all PCB congeners and failed implantation wasn’t significant. Finally, the researchers considered no other confounding risk factors for the failed implantation. No wonder this was published in EHP.