Choosing the Best Data for the Analysis at Hand

There are times when there is more than one data set that can be used to conduct the analysis that the researcher is focused on. Other times there is only one source of data.

Today I was surprised when I visited the Census Bureau website and read their “Stat of the Day.” The title was “Males Per Females.” The text read:

According to the American Community Survey, in 2010 the national average was 96.7 males per 100 females. The states of Alaska (109.0), North Dakota (103.7), and Wyoming (103.5) had the highest ratio.

The American Community Survey is a very large survey conducted by the Census Bureau. The primary goal of the survey is to provide estimates at very low geographic levels. But this “Stat of the Day” focuses on national and state data. What puzzles me is that if giving such statistics for 2010 would not the Decennial Census tabulations be a much preferred source for these numbers. This is even more surprising as those same Decennial Census numbers would be the ones used in the weighting of the data in the American Community Survey?

I just don’t get it………….