Frequently Asked Questions
What procedures guide Illinois BRFSS data collection?
The document http://www.cdc.gov/brfss/data_documentation/PDF/UserguideJune2013.pdf
has detailed information required of Illinois and other states participating in the BRFSS; including survey protocol, questionnaire development, and survey methodology.
How does Illinois collect the data?
Using a scientifically designed method, disproportionate stratified sampling (DSS), a random sample
of telephone numbers is selected each month for telephone interviews. Beginning in 2011, both
landline and cellular telephone numbers were included in the sample. The surveys are conducted
year round on a monthly basis. The Illinois county surveys (ICBRFS) include the same core topics as
the statewide, but the sample selection is by county area. Procedures used to conduct interviews are
the same for all surveys. Interviewers follow carefully controlled procedures to assure that the
collected data meet standards of acceptability. Once all the data for a year are collected, the dataset
is edited in preparation for weighting.
How is the Illinois Behavioral Risk Factor data presented?
Data-users may select frequency or cross-tabulation data for any of three categories: Illinois, strata, or
county. Within the strata, any of five strata may be selected: Chicago; Suburban Cook County; collar
counties; urban counties; or rural counties. The collar counties include DuPage, Kane, Lake, McHenry
and Will. Urban counties include Champaign, DeKalb, Kankakee, Kendall, McLean, Macon, Madison,
Peoria, Rock Island, Sangamon, St. Clair, Tazewell and Winnebago. The remaining 83 counties are
grouped as rural counties.
What is weighting?
Weighting is a process that attempts to remove bias in the sample. Prior to 2011, post stratification
was used to weight the data using age, sex, race, ethnicity and region. Beginning in 2011, the raking
method (design weighting and iterative proportional fitting) was employed and added telephone
ownership, education level, marital status, and home ownership to the variables needed to weight the
Why should data users be concerned about how the data is weighted?
During the years 1998 through 2003, a split sample was used to collect data in Illinois. The Illinois
BRFSS program used a dual questionnaire procedure to collect data on a variety of health-related
subjects. This sample was split in two with each half being asked different versions of the survey. Both
Illinois versions used the same core questions as the rest of the states, but different state-added
questions were asked of each group. An important point to note is that the Illinois BRFSS data were
weighted with a process that took the elements of the split sample and dual questionnaire into
consideration. By excluding the split sample factor, other sources may not present 1998 through 2003
Illinois BRFSS data with the proper weighting factors.
Additionally, 2011 marked the year a new raking weighting formula was instituted and cell phones
were added to the sampling frame. Because of this, data collected prior to 2011 is not comparable to
data collected from 2011 forward.
For more information about the BRFSS program at CDC, visit that agency's website at www.cdc.gov/brfss.