Surveys Used in the Analysis

This study employs three years of data collected from the 1997 (black sample n=1,269), 1998 (n=996), and 1999 (n=1,001), Gallup Poll Social Audits on Black/White Relations, also known as the Gallup Race Relations Social Audits (GRRSA). For these surveys, Random Digit Dialing (RDD) telephone interviews were carried out in each of the three years, with each year containing an oversample of blacks. Samples of telephone numbers from Survey Sampling Inc. were used to conduct the survey. The RDD samples used stratified probability designs to represent the adult population living in households with telephones within the continental United States. The GRRSAs were systematically designed to update long-term Gallup Poll trends on race relations, and used a rigorous methodology with sufficiently large sample sizes for in-depth subgroup analysis. While all three years included more than one racial ethnic group, the analyses here were only performed on the black respondents. For all three years the sampling error was ± 5 percent.