USING FOUND DATA TO AUGMENT A PROBABILITY SAMPLE - PROCEDURE AND CASE-STUDY

While probability sampling has the advantage of permitting unbiased po pulation estimates, many past and existing monitoring schemes do not e mploy probability sampling. We describe and demonstrate a general proc edure for augmenting an existing probability sample with data from non probability-based surveys ('found' data): The procedure, first propose d by Overton (1990), uses sampling frame attributes to group the proba bility and found samples into similar subsets. Subsequently, this simi larity is assumed to reflect the representativeness of the found sampl e for the matching subpopulation. Two methods of establishing similari ty and producing estimates are described: pseudo-random and calibratio n. The pseudo-random method is used when the found sample can contribu te additional information on variables already measured for the probab ility sample, thus increasing the effective sample size. The calibrati on method is used when the found sample contributes information that i s unique to the found observations. For either approach, the found sam ple data yield observations that are treated as a probability sample, and population estimates are made according to a probability estimatio n protocol. To demonstrate these approaches, we applied them to found and probability samples of stream discharge data for the southeastern US.