People who approach an issue with certain beliefs are generally less likely to check claims that comport with those beliefs. It's called confirmation bias. Observe: Tuesday's New York Times carried this correction, highlighted by John Hinderaker at Powerline:
An article on May 7 about the Obama administration's appointment of a panel of experts to find ways to make hydraulic fracturing safer misstated the prevalence of cases in which fluids from the gas drilling process have been proven to have contaminated drinking water. There are few documented cases, not numerous ones, although federal and state investigations into reports of such incidents are continuing.
In other words, hydraulic fracturing is not, by and large, a danger to drinking water supplies. Since potential dangers to drinking water are integral to virtually every argument mounted against the practice, the incidence of contamination is crucial to the debate.