6.1 Comments on the selection of variables for analysis
We use the weighting variables, such as, sex, age group (45-48, 49-54, 55-64, 65-74, and 75-85), education level (low, medium, high lower, and high upper), to calibrate the survey weights. It is recommended that these weighting variables should be included in the initial model. The stratum indicator variables, i.e., the provinces for the Tracking cohort or the provinces crossed with individual DCS areas for the Comprehensive cohort and the pooled cohorts, may also be included since they were also used in calibration of the weights. Some of these variables may be removed in the final model if it is shown that they are not significant.
6.1 Comments on the selection of variables for analysis
We use the weighting variables, such as, sex, age group (45-48, 49-54, 55-64, 65-74, and 75-85), education level (low, medium, high lower, and high upper), to calibrate the survey weights. It is recommended that these weighting variables should be included in the initial model. The stratum indicator variables, i.e., the provinces for the Tracking cohort or the provinces crossed with individual DCS areas for the Comprehensive cohort and the pooled cohorts, may also be included since they were also used in calibration of the weights. Some of these variables may be removed in the final model if it is shown that they are not significant.