Technical Information and Data
BRFSS POLICY MEMO 97.6
CDC Processing and Publication of Data
BRFSS Policy Memo 94.3 describes criteria for acceptable sampling designs
to insure that data are obtained from a random sample of states'
populations. Additionally, it has long been a requirement that states use
wording of questions on the Core, Rotating Core, Emerging Core, and Optional
Modules without modification (BRFSS Policy Memo 93.3).
Increasingly, BRFSS data are being utilized by a variety of groups. The
Behavioral Surveillance Branch (BSB) is fully supportive of more widespread
distribution and use of these data. One of the most important
responsibilities of the branch is to ensure that BRFSS data are collected in
a manner that enables comparison across states. Thus, it is essential that
samples are scientifically drawn in all participating states, and that no
changes (either in wording or sequence) are made to the questions provided
by the branch on the core, rotating core, and optional modules.
Based upon recommendations from the American Statistical Association (ASA)
BRFSS Survey Methods working group, data from states that do not conform to
Memo 94.3 may not be included in BSB publications such as the annual BRFSS
Summary Prevalence Report and MMWR Surveillance Summaries, and such data may
not be included in the final CDC main-frame BRFSS computer file or other
electronic formats such as CD-ROM (i.e., they may not be made available to
researchers at CDC or else-where for analysis). However, data from states
that use nonprobability sampling, that do not provide information on each
record that reflect the sample design, or that do not include all possible
household numbers in its sampling frame will continue to be processed,
cleaned, edited, and weighted (if appropriate) and returned to states that
submit them. Because of the changing nature of survey sampling, BSB staff
annually will review the state of the art on acceptable sampling methods and
notify states as changes occur.
States must provide BSB written copies of their questionnaires by
February 15th of each data collection year, and notify BSB if any changes
are made to questionnaires during the course of the year. Core, Rotating
Core, Emerging Core, and Optional Module questions must not be modified by
states. Again, based upon recommendations from the American Statistical
Association (ASA) working group, if such modifications occur, data on
modified questions will not be included in BSB publications such as annual
BRFSS Summary Prevalence Report and MMWR Surveillance Summaries, nor will
data from these questions be included on the CDC mainframe BRFSS computer
file or other electronic formats such as CD-ROM (i.e., they will not be made
available researchers at CDC or elsewhere for analysis).