interpretation of the weighted average using WT06


I’d like to estimate the total hours an adult spent doing exercise, work, travel…etc over a three month period of time (not just one day) using a single year data (e.g. 2010). The description of WT06 stated that this variable is the “preferred one in most cases” and the construction of WT06 ensures that“…the sums of the respondent weights add up to the appropriate number of weekday person-days and weekend person-days over the quarter”.

The latter statement is not totally clear to me….so if I set WT06 as the weighting variable in Stata, would the estimated average be (1) the hours spent working/doing exercise…. in a day, although the estimates represent an ‘average day’ (could be weekend or a weekday) over a quarter, or simply (2) the total time spent doing work/sport…etc. over three months?

If it is (1), does it make sense to multiply the daily average by 91 to get the total hours of time use over a three-month period of time?



If you set WT06 as the weighting variable in STATA, the estimated average would be for the total time a person spent doing an activity over a 3-month (calendar quarter) period. The weighted average does not represent a single day; therefore, you should not multiply the average by 91. Keep in mind that you should use pweight in STATA.

Hope this helps.