Download:
pdf |
pdfBehavioral Risk Factor Surveillance System
Weighting BRFSS Data
BRFSS 2013
Weighting BRFSS Data
Introduction: Weighting Rationale
When data are used without weights, each record counts the same as any other record—and this implies
the following assumptions:
1) that each record has equal probabilities of being selected, and
2) noncoverage and nonresponse are equal among all segments of the population.
During data collection, the conditions that actually happen may be different than the above assumptions,
and the differences can become large enough to affect the results of the data set. Weighting each record
appropriately can help to adjust for these differences (or “assumption violations”). Another reason for
weighting—one that is conceptually unrelated to the first reason--is to make the total number of cases
equal to some desired number which, for state BRFSS data, is the number of people in the state who are
aged 18 years and older. In the BRFSS, such raking serves as a blanket adjustment for noncoverage and
nonresponse and forces the total number of cases to equal population estimates for each geographic
region, which for the BRFSS sums to the state population.
This document provides a general description of the process that reflects factors taken into
account in weighting the 2013 BRFSS data. Where a factor does not apply, its value is set to one for
calculation. Regardless of state sample design, use of the final weight in analysis is necessary if users
are to make generalizations from the sample to the population.
Weighting Methods and BRFSS
From 1984 to 2010, BRFSS used a technique called post stratification to weight BRFSS survey data to
account for known proportions of age, race and ethnicity, gender, geographic region, and other known
characteristics of a population. This type of weighting was important because it helped the sample
represent the population accurately and adjusted for nonresponse bias. In 2011 a new statistical method
called raking replaced the older method because it could help record additional population
characteristics such as education level, marital status, and home ownership status of respondents. It also
helps the data fit into an analysis-friendly format called dual-frame survey design, so organizers could
account for the type of telephones the respondents used while completing the survey. Because of the
changes in the methodology, researchers are advised to avoid comparing data collected before the
changes (up to 2010) with data collected from 2011 and onward. Details of these methodology changes
can be found with the BRFSS 2011 annual data page:
http://wwwdev.cdc.gov/brfss/annual_data/annual_2011.htm
More About Raking
The raking weighting methodology is comprised of two sections: design weight and raking.
Design Weight = STRWT * (1/NUMPHON2) * NUMADULT
In 2013, the inclusion of cellular telephone respondents who received between 90 and 99 percent of their
telephone calls on their cellular telephone required an adjustment to the design weights to account for
the overlapping sample frames. From each of the two sample frames, BRFSS calculated a compositing
factor for the mostly cellular telephone dual sampling frame users. BRFSS did this calculation by
multiplying the design weight by the compositing factor to generate a composite weight for the records
in the overlapping sample frames. BRFSS then truncated the design weight based on quartiles within
geographic region, which BRFSS uses as the raking input weight. To increase the value of extremely
low weights and decrease the value of extremely high weights (and to reduce errors in the outcome
estimates that could result), BRFSS uses weight trimming. The objective of weight trimming is to
reduce errors in the outcome estimates caused by unusually high or low weights in some categories.
The stratum weight accounts for differences in the basic probability of selection among strata (subsets of
area code/prefix combinations). It is the inverse of the sampling fraction of each stratum. There is rarely
a complete correspondence between strata, which BRFSS defines by subsets of area code/prefix
combinations, and regions, which BRFSS defines by the boundaries of government entities.
The stratum weight (STRWT) is calculated using the following items:
•
•
•
Number of available records (NRECSTR) and the number of records selected (NRECSEL)
within each geographic strata and density strata.
Geographic strata (GEOSTR), which may be the entire state or a geographic subset such
as counties, census tracts, etc.
Density strata (_DENSTR) indicating the density of the phone numbers for a given block
of numbers as listed or not listed.
Within each _GEOSTR*_DENSTR combination, the stratum weight (_STRWT) is calculated from the
average of the NRECSTR and the sum of all sample records used to produce the NRECSEL. The stratum
weight is equal to NRECSTR / NRECSEL.
1/ NUMPHON2
The inverse of the number of residential telephone numbers in the
respondent’s household.
NUMADULT
The number of adults 18 years and older in the respondent’s household.
FINAL WEIGHT
BRFSS rakes the design weight to 8 margins (age group by gender,
race/ethnicity, education, marital status, tenure, gender by race/ethnicity,
age group by race/ethnicity, phone ownership). If BRFSS includes
geographic regions, four additional margins (region, region by age group,
region by gender, region by race/ethnicity) are included. If at least one
county has 500 or more respondents, BRFSS includes four additional
margins (county, county by age group, county by gender, county by
race/ethnicity).
_LLCPWT
The final weight assigned to each respondent.
BRFSS uses weight trimming to increase the value of extremely low weights and decrease the value of
extremely high weights. The objective of weight trimming is to reduce errors in the outcome estimates
caused by unusually high or low weights in some categories.
Calculation of a Child Weight
BRFSS calculates the design weight for the child weighting from the stratum weight times the inverse of
the number of telephones in the household and then multiplies it by the number of children:
Child Design Weight = STRWT * (1/NUMPHON2) * CHILDREN
CHIILDWT =
BRFSS rakes the child design weight is raked to 5 margins including age by
gender, race/ethnicity, gender by race/ethnicity, age by race/ethnicity, and phone ownership.
_CLLCPWT is the weight assigned for each child interview.
Calculation of a Household Weight
BRFSS calculates the design weight for the household weighting from the stratum weight times the
inverse of the number of telephones in the household.
Household Design Weight = STRWT * (1/NUMPHON2)
to the Number of Persons in a Household,
Number of Adults in a Household, and Number of Children in a Household
BRFSS rakes HOUSEWT = STRWT * (1/ NUMPHON2)
_HHOLDWT is
the weight assigned to each household.
BRFSS Weighting prior to 2011
Please note prior to 2011, the data weighting formula used post-stratification as the weighting method.
The formulas listed below apply to data BRFSS weighted prior to 2011.
FINALWT = STRWT * 1 OVER NPH * NAD * POSTSTRAT
BRFSS intended the computational formula above to reflect all the possible factors that it could take into
account in weighting a state’s data. Where a factor does not apply, BRFSS set its value to one.
FINALWT is
the final weight BRFSS assigns to each respondent.
The stratum weight accounts for differences in the basic probability of selection among strata (subsets of
area code/prefix combinations). It is the inverse of the sampling fraction of each stratum. There is rarely
a complete correspondence between strata, which BRFSS defines by subsets of area code/prefix
combinations, and regions, which are defined by the boundaries of government entities.
BRFSS calculates the stratum weight (STRWT) using the following items:
•
•
•
Number of available records (NRECSTR) and the number of records selected (NRECSEL)
within each geographic strata and density strata.
Geographic strata (GEOSTR), which may be the entire state or a geographic subset such
as counties, census tracts, etc.
Density strata (_DENSTR) indicating the density of the phone numbers for a given block
of numbers as listed or not listed.
Within each _GEOSTR*_DENSTR combination, BRFSS calculates the stratum weight (_STRWT) from
the average of the NRECSTR and the sum of all sample records used to produce the NRECSEL. The
stratum weight is equal to NRECSTR / NRECSEL.
The inverse of the number of residential telephone numbers in the respondent’s
household.
1/ NPH
NAD
The number of adults 18 years and older in the respondent’s household.
POSTSTRAT is
the number of people in an age-by-gender or age-by-race-by-gender category in
the population of a region or a state divided by the sum of the products of the preceding weights
for the respondents in that same age-by-gender or age-by-race-by-gender category. It adjusts for
non-coverage and non-response and, before 1995, also adjusts for different probabilities of
selection by region, where applicable.
Calculation of a Child Weight prior to 2011
CHILDWT = STRWT * 1 OVER NPH * CHILDREN * POSTCH
BRFSS intends the computational formula above to reflect all the possible factors that could be taken
into account in weighting a state’s data. Where a factor does not apply, BRFSS sets its value to one.
CHILDWT is
the final weight BRFSS assigns to each randomly selected child as identified by the adult
respondent.
BRFSS calculates the design weight for the child weighting from the stratum weight times the inverse of
the number of telephones in the household and then multiplies it by the number of children:
Child Design Weight = STRWT * (1/NPH) * CHILDREN
POSTCH is the number of children in an age-by-gender or age-by-race-by-gender category in the
population of a region or a state divided by the sum of the products of the preceding weights for
the children in that same age-by-gender or age-by-race-by-gender category. It adjusts for noncoverage and non-response.
Calculation of a Household Weight prior to 2011
HOUSEWT = STRWT * 1 OVER NPH * POSTHH
HOUSEWT is
the weight BRFSS assigns to each household.
BRFSS calculates the design weight for the household weighting from the stratum weight times the
inverse of the number of telephones in the household:
Household Design Weight = STRWT * (1/NPH)
POSTHH is
the number of households in the population of a region or a state divided by the sum
of the products of the preceding weights for the households in that same category. It adjusts for
non-coverage and non-response.
File Type | application/pdf |
File Title | BRFSS: Weighting BRFSS Data 2013 |
Subject | BRFSS: Weighting BRFSS Data 2013 |
Author | CDC - BRFSS |
File Modified | 2014-08-18 |
File Created | 2014-08-18 |