3 ms·
I got it by eyeballing Table 10 to get about 5.5 hours worked/day, multiplied by 7. Assuming the sample is statistically representative, the hours/week calcula
by yummyfajitas 15y ago
I got it by eyeballing Table 10 to get about 5.5 hours worked/day, multiplied by 7.
Assuming the sample is statistically representative, the hours/week calculation is accurate. Across the population as a whole, it is a mathematical identity that:
Sum(Sum(hours(day, teacher), all teachers), days in week) == Sum(Sum(hours(day, teacher), days in week), all teachers)
In a statistical sample, the equality will be approximate due to sampling error, and it's worthy of a note. Maybe the ATUS did a bad job, but in general I assume they do an adequate job absent evidence they did not. Do you have better data?