UC San Diego

Random Digit Dial Telephone Surveys (RDDTSs) are often used in public health and psychology research. Data from such surveys suggest that alcohol consumption rates are similar for African American (AA) and Caucasian adults in California (CA). While these finding are reliable, their validity is questionable because of the methodological problems present in RDDTS. Data collected from RDDTS are gathered from non-representative samples of higher SES landline phone owners and AA women residents of integrated neighborhoods. Thus, the validity of findings from RDDTS data is questionable. The purpose of the current study was to overcome the methodological problems that are inherent in RDDTS by examining the prevalence of binge drinking in a random, statewide sample of 2,190 CA AA adults, using a community-based participatory research approach, and to examine possible sociocultural variables that may contribute to alcohol use. Participants were 54% female, 46% male AAs, with ages ranging from 18 to 95. The results suggested that the current Community-Based Sample (CBS) was more representative of the CA AA population than CA RDDTS samples. Specifically, it was more diverse in age, was younger, and had a larger percentage of AA men and low -income adults. In addition, the current community-based sample reported engaging in significantly more binge drinking behavior than AA and Caucasian CA RDDTS samples, and that binge drinking was even more prevalent among African Americans who lacked landline telephones. Finally, multilevel statistical modeling revealed that neighborhood SES was a significant predictor of binge drinking, but neighborhood segregation was not. Cross-level interactions between measures of individual-level SES and neighborhood SES and individual SES and segregation were not significant. Moreover, social (e.g., racial discrimination, neighborhood dangerousness) and cultural (e.g., acculturation) factors did not significantly predict binge drinking behavior. Findings from the current study suggest that using a community-based participatory research approach may circumvent some of the methodological problems inherent in RDDTS and result higher quality alcohol behavior data