Hispanic Community Health Study / Study of Latinos

This variable is an indicator of household sample address allocation and distribution for HCHS/SOL recruitment.

This variable is the sample design list household frame count variable used in statistical analyses.

This variable is the sample design variable used in statistical analyses.

This variable is the sample design clustering household frame count variable used in statistical analyses.

This variable is the sample design clustering variable used in statistical analyses as the cluster variable that is combination of the field center and selected block group identifier.

This variable is the sample design stratification variable used in statistical analyses as the stratum variable that is a combination of field center, Hispanic household proportion, and SES.

This variable is the sample design stratification household frame variable used in statistical analyses.

The calibrated sample weight is the calibrated, trimmed, non-response adjusted reciprocal of a participant's probability of selection into the HCHS/SOL study. The trimmed, nonresponse adjusted sample weights were calibrated using first the 7-category Hispanic Background distribution and then the 12-category age group/gender distribution using the 2010 Decennial Census. This variable is used for analyses purposes only.

The FC-specific normalized sample weight is the field center specific normalized, calibrated, trimmed, non-response adjusted (household and person-level) reciprocal of a participant's probability of selection into the HCHS/SOL study. The calibrated, trimmed, non-response adjusted sample weights were normalized using the weighted mean divided by weighted standard deviation calculated for each field center separately and used to normalize the expanded calibrated sample weights (WEIGHT_Final_Norm_Center). This final sample weight variable is used for analyses purposes only when the subpopulation is restricted to within a single site.

The normalized sample weight is the overall normalized, calibrated, trimmed, nonresponse adjusted (household and person-level) reciprocal of a participant's probability of selection into the HCHS/SOL study. The base sample weights, i.e. the reciprocal of a participant's probability of selection into the HCHS/SOL study, were first adjusted by household-level and person-level non-response. The non-response adjusted sample weights were then trimmed using the selected average percentile that minimizes model MSE within each field center. All sample weights that fell above the selected percentile were trimmed to the selected percentile and the trimmed difference was evenly distributed among all of the non-trimmed sample weights. The trimmed, non-response adjusted sample weights were then calibrated using first the 7-category Hispanic Background distribution and then the 12-category age group/gender distribution using the 2010 Decennial Census. Finally, these calibrated, trimmed, non-response adjusted sample weights were normalized using the weighted mean divided by weighted standard deviation of the entire HCHS/SOL sample to normalize the expanded calibrated sample weights (WEIGHT_Final_Norm_Overall). This final sample weight variable is used for analyses purposes only.