Adjusted β coefficients and 95% confidence limits (CLs)* for the association between survey design elements and prevalence differences between 2000 Behavioral Risk Factor Surveillance System (BRFSS) and 2000 Census for factors included and not included in the adjustment weight for non-response/non-coverage (NR/NC).
| Factors included in adjustment weight for NR/NC | Factors not included in adjustment weight for NR/NC | |||||
| Survey design elements | % Women adjustment β (95% CL) | % White adjustment β (95% CL) | % 18–34 y adjustment β (95% CL) | % Hispanic adjustment β (95% CL) | % Income <$15000 adjustment β (95% CL) | % Married adjustment β (95% CL) |
| State level† | ||||||
| BRFSS response rate | 0.00 (−0.01 to 0.02) | −0.02 (−0.11 to 0.07) | 0.01 (−0.05 to 0.07) | −0.04 (−0.14 to 0.07) | 0.01 (−0.08 to 0.09) | 0.01 (−0.06 to 0.09) |
| In-house data collection (vs contract) | −0.17 (−0.57 to 0.22) | 3.06 (0.14 to 5.98) | −0.94 (−2.77 to 0.89) | 0.24 (−1.58 to 2.06) | −0.22 (−2.23 to 0.41) | 0.61 (−0.73 to 1.96) |
| Sampling fraction | −0.60 (−1.36 to 0.15) | 0.76 (−3.66 to 5.18) | 1.67 (−1.97 to 5.30) | 1.42 (−1.42 to 4.25) | −3.42 (−6.79 to −0.05) | 0.84 (−1.49 to 3.18) |
| Number of call attempts | −0.29 (−0.48 to −0.10) | 0.98 (−0.23 to 2.19) | −0.29 (−1.09 to 0.50) | 0.04 (−1.15 to 1.24) | 0.40 (−0.41 to 1.21) | 0.32 (−0.61 to 1.25) |
| County level† | ||||||
| BRFSS CASRO response rate | 0.01 (−0.01 to 0.03) | −0.02 (−0.03 to 0.01) | 0.01 (−0.01 to 0.03) | 0.01 (0.00 to 0.03) | −0.02 (−0.05 to 0.01) | −0.01 (−0.04 to 0.02) |
| In house data collection (vs contract) | −0.70 (−1.71 to 0.32) | 3.13 (0.25 to 6.01) | −0.63 (−1.76 to 0.49) | 0.81 (−0.58 to 2.20) | 0.11 (−2.11 to 2.32) | −0.06 (−1.47 to 1.35) |
| Sampling fraction | −0.29 (−1.88 to 1.30) | 1.10 (−2.52 to 4.73) | 1.19 (−1.24 to 3.63) | 1.35 (−0.72 to 3.42) | −2.96 (−7.16 to 1.24) | −0.81 (−3.40 to 1.77) |
| Number of call attempts | −0.25 (−0.89 to 0.38) | 0.59 (−0.63 to 1.81) | −0.45 (−1.31 to 0.41) | 0.63 (−0.01 to 1.27) | 0.56 (−0.46 to 1.58) | −0.09 (−0.86 to 0.68) |
↵* Robust regression used to estimate β coefficients and 95% confidence limits controlling for Census region and telephone coverage (results not reported).
↵† State level analyses included 50 states and DC. Number of counties for county level analyses varied due to missing data and minimum sample size requirements (gender: n=732; race: n=730; age: n=732; ethnicity: n=729; income: n=638; marital status: n=730).