Once the residences were selected, lists of residents and their
addresses and telephone numbers in the 72 sampled counties were
developed. For each residential location, the list included 50
essentially equivalent residences to provide alternate participants
if residents of the selected locations declined to participate.
Ultimately, participants at 144 residential locations were recruited
for the study. Fig. 1 shows the counties from which the samples
were obtained.
The first of two water collections at each location were made
during the winter months, and the second in the spring. This allows
the estimation of variance due to date of collection as well as that
due to location within county. In nine locations, the initial
participant was unable to provide a water sample for the second
collection and a nearby neighbor was recruited to provide the
second sample. Agents were sent to each residence with two
250 mL high density polyethylene bottles and a survey form
requesting information on water source (municipal, well), pipes
and use of water softeners and treatments. The participants were
instructed to collect the water from the kitchen tap and to let the
water run for 3 min before filling the two bottles. The filled bottles
were immediately shipped by the agents via ground service at
ambient temperature to the Food Analysis Laboratory Control
Center (FALCC) at Virginia Tech (Blacksburg, VA) for sample
preparation. The water samples were aliquoted into small bottles
and stored at 60 8C prior to analysis.