WT03 is a 15-digit numeric variable reporting the ATUS respondent probability weights for 2003 only.
WT03 provides ATUS respondent weights for 2003 only based on the 2003 weighting methodology. Under the method used in 2003, the sum of the respondent weights equals the number of person-days in the month for the population as a whole and the weighted totals for the month correspond to the number of weekdays, Saturdays and Sundays over the month.
All ATUS weights are probability weights. Users should take care to specify them properly in their statistical packages.
The methodology used to construct respondent weights has been updated in each year of the ATUS. In 2004, the weighting methodology was changed to ensure that weights would sum to the number of person-days in the month for selected subpopulations as well as for the population as a whole, though not necessarily that the shares of weekdays, Saturdays and Sundays in the weighted data would equal the actual shares for the selected subpopulations. The 2005 methodology produces weights that the Bureau of Labor Statistics believes can be treated as comparable to those produced using the 2004 methodology. This methodology produces weights that sum to the number of person-days over the quarter, rather than over the month, and weighted totals match the number of weekday and weekend person-days, rather than the number of weekday, Saturday and Sunday person-days. The variable WT04, available for 2003 through 2005, was constructed for 2003 and 2004 using the 2004 methodology and for 2005 using the 2005 methodology.
WT06 is the recommended weighting variable for most analyses. The 2006 weighting methodology ensures that the sums of the respondent weights add up to the number of person-days on weekdays and on weekend days over the quarter, both for the population as a whole and for selected subpopulations. The variable WT06, available for all years, was constructed using the methodology introduced in 2006. This is the only weighting variable that is suitable for use in analyses that combine time use data from 2006 with data for other years.
- 2003: All ATUS respondents.