This paper explores various approaches to achieving representative and
generalizable practice-based data. The data used were derived from or
iginal research conducted by the Pediatric Practice Research Group (PP
RG), a research consortium including full-time staff at Children's Mem
orial Hospital and 132 practitioners in 35 Chicago-area practices in a
variety of settings ranging from Chicago's inner city to rural northw
est Indiana. Sampling issues addressed include sampling frame, source
(population or practice), geographical representation, seasonal repres
entation, survey instrument design, socioeconomic diversity, and conta
ct and response rates.