Methodology

Target Population

Adults 18 years or older who live in a private residence or a college housing unit are asked to take part in the survey. Only one adult per household is randomly selected to be interviewed once the telephone number is verified as a private working phone number.

Data Collection

Information is collected through telephone interviews (cell phone and landline) throughout the year. A set of randomly generated phone numbers are produced at least four times per year, and the phone numbers are first called to verify if they are residential or a business phone number. If residential, a randomly selected adult in the household is asked to participate. This can result in repeat phone calls to the household in order to get the appropriate (randomly selected) adult to participate.

Sample Size

Since 2003, the goal of the Virginia BRFSS has been to complete at least 5,000 interviews with Virginia adults annually.

Cell Phone Sampling

Virginia was one of six states that participated in a cell phone sample pilot in 2008. The goal was to complete at least 250 BRFSS interviews with respondents on their cell phones. For participating and completing an interview, respondents were given a ten dollar ($10) Amazon gift card. In 2009 and 2010, Virginia continued to complete at least 250 BRFSS interviews with respondents on their cell phones. The cell phone data from 2008-2010 were not included in the yearly BRFSS datasets. Cell phone data was first included in the BRFSS yearly dataset in 2011.

Stratification

From 1997 to 1999, and again from 2002 through 2010, the Virginia Department of Health (VDH) funded additional data collection to allow stratification of the state by the thirty-five (35) health districts. Approximately 140 interviews were collected annually in each health district.

Link to the thirty-five (35) local health districts websites

Weighting Methodology

BRFSS data are first weighted for the probability of selection of a telephone number, the number of adults in a household, and the number of telephones in a household. Then post-stratification adjustments are made to account for non-response and non-coverage (households without telephones). For the 2010 dataset and prior surveys post-stratification was based on only age, sex, and region of residence. Beginning with the 2011 dataset, a raking (or iterative proportional fitting) procedure is used for post-stratification adjustment. The use of raking procedures allows post-stratification adjustment to be based on twelve demographic dimensions rather than only three. Including new variables in the weighting process can reduce the potential for selection bias while increasing representation. For a complete explanation of raking procedures.