A DIFFERENT TYPE OF SELF-SERVING SAMPLING BIAS

Regarding the Lanxess Canada site in Elmira, Ontario I have previously elaborated on locational sampling biases. These have included much greater sampling at locations along the creek that have public road access as well as once Chemtura/Lanxess found "hot spots" they focused all their future sampling on these locations. What that did was to artificially reduce the likelihood of finding other "hot spots" downstream from the former Uniroyal Chemical site. ...................................................................................................................... Today's sampling bias is even more insidious. There is an old saying that "You will never find what you don't look for." In other words if you don't want to find a lot of mercury in the Canagagigue Creek than don't test for it. If you are particularly worried about finding PCBs in either soils or sediments then for gosh sakes don't analyse in the lab for PCBs. The very same for Polycyclic Aromatic Hydrocarbons (PAH). The same strategy goes for pesticides other than DDT. If you want to avoid finding Endrin, endosulfan, dieldrin, lindane and so many others then avoid testing for them in the lab. Problem solved. You see if there is little transparency combined with minimal accountability then you the polluter (and friends) are off to the races. ................................................................................................................ I'm currently going through Appendix C in the recently released Draft Risk Assessment (HHERA) for the Canagagigue Creek. While there are a huge number of sediment, soil and creekbank samples there is a dearth of compounds being tested for. It's all about dioxin/furans and DDT with a few solvents, metals, PCBs and PAHs occasionally thrown in for filler. If you test for PCBs ten times over ten years then the best you will ever get on average is one detection per year at a 100% detection rate. Imagine however if you test for DDT one hundred times a year for ten years! If you only had a detection rate of 25% you would still average 25 detections per year. This is the joy and the shame of dishonest statisticians. This is how polluters and fellow travellors deceive citizens and remove other present toxic compounds from consideration.