This book is based on a variety of information sources, but the primary data used here come from a national survey conducted by the authors between October and December, 2002 of 1,791 adults 18 years of age or older (619 whites, 565 blacks, and 607 Hispanics). Our sample also included 310 youths 13–17 years old, about 100 from each racial group. Although we occasionally quote some of our young interviewees, we do not systematically analyze the youth data due to small sample sizes. Only respondents who reside in metropolitan areas of at least 100,000 population (either cities or adjacent suburbs) were included in our sample because it is in such locales that policing is likely to be especially salient and, perhaps, contentious for residents – unlike in more rural areas (Weisheit, Falcone, and Wells 1995). Our sample is representative of citizens living in telephone households in areas that meet our population size criterion.

SAMPLING

The survey was conducted for the authors by Knowledge Networks, Inc., a Web-based survey research firm that combines probability sampling with the reach and capabilities of the Internet to yield representative samples of respondents without sacrificing data quality. Research comparing the quality of data yielded by Knowledge Networks' Web-based survey methodology with that of random digit dialed (RDD) telephone surveys has found that Knowledge Networks yields representative samples that produce parameter estimates very similar to the estimates of RDD samples (Krosnick and Chang 2001; Baker et al. 2003; Berrens 2003).