Wisconsin Behavioral Risk Factor Survey (BRFS)
Wisconsin BRFS is part of the Centers for Disease Control and Prevention's (CDC) Behavioral Risk Factor Surveillance System (BRFSS), which comprises separate health surveys in all 50 states and several U.S. territories. The Wisconsin BRFS Program is located in the Office of Health Informatics, Division of Public Health. Wisconsin BRFS telephone interviews are conducted by the University of Wisconsin Survey Center. See the CDC BRFSS website.
Sample Design and Protocol
As of 2011, BRFS includes both landline and cell phone interviews, obtained from separate samples and later combined and weighted by CDC. See the BRFSS User Guide (PDF, 204 KB) for more information about combined landline/cell phone data and other system-level BRFS topics.
Wisconsin BRFS uses disproportionate stratified sampling based on Department of Health Services regions, county population sizes and, in some instances, the demographic composition of geographic areas. CDC weights annual data files to represent the Wisconsin adult population ages 18-plus living in households with a landline or cell telephone, or both. The landline sample excludes adults living in institutions and other group quarters. The cell phone sample also excludes people living in institutions, but allows for the inclusion of college students living in dormitories. Few such cases are in the final set of completed interviews, however.
Within each sample stratum identified by state BRFS programs, CDC’s sampling contractor generates batches of randomly selected telephone numbers and provides them to BRFS data collection contractors on a quarterly basis. One adult is randomly selected to complete the interview in each eligible landline sample household. BRFS cell phone interviews are conducted with the adults reached at sampled numbers, and no household screening takes place. Adults reached by cell phone who also have a landline phone are interviewed without regard to the proportion of calls received by cell phone. Dual phone use is accounted for in post-survey weighting. Data from completed cell phone interviews with respondents whose area code and state of residence are discordant are assigned to the state of residence.
Weighting of annual BRFS data files is done by CDC for all states and territories. Data are initially weighted to correct for disproportionate sampling, multiple phone numbers to the household (landline), and dual landline-cell phone use. Following this, case weights for data from 2011 and later are calculated by CDC using iterative proportional fitting, or "raking." The raking procedure uses demographic information from the U.S. Census Bureau's American Community Survey, and several demographic variables in BRFS data files, to create the final weights. The weighting variable for each year of data, when used in analysis, produces estimates that are representative of the state’s adult population.
IMPORTANT: Estimating Population Numbers with Multiple Years of Data
Each year's BRFS weights sum to the total adult population. To estimate population numbers using more than one year of data, divide "Estimated Number in the Population" shown in the results by the number of years to resolve the inflated numbers generated by multiple years of summed weights. This will produce an average estimated population number for the time period.
For example, selecting General Health Status for the years 2003-2005 produces an estimated 7,198,600 adults who were in excellent or very good health – an inflated number. Dividing this result by three, the number of data years, yields 2,399,533, which rounds to approximately 2,399,000, or 2.4 million, adults in excellent or very good health.