Neighborhood data from the American Community Survey (ACS) comes from a survey mailed to a small percentage of households in each neighborhood.
The margin of error (MOE) is an indicator of the reliability of ACS estimates. Adding the MOE to the estimate provides an upper limit and subtracting the MOE from the estimate provides a lower limit of the range where the true value of the estimate most likely actually falls.
Here are some examples of how you can write about this data in a grant report:
“From 2010 to 2014, somewhere between 59.1% and 59.9% of people commuted less than 30 minutes to work.”
“From 2010 to 2014, at least 20% (or no more than 30%) of people in a neighborhood live below the poverty line.”
“From 2010 to 2014 the Census estimates that 30.6% of people traveled between 30 and 60 minutes, although this percentage could range from 30.3% to 30.9%.”
The margin of error (MOE) makes it tricky to compare different places or timeframes. For instance, it is hard to tell if a poverty rate of 10% (+/–2%) is really higher than a poverty rate of 7% (+/– 2%) even though the two estimates are different.
The widget below will do a calculation for you and let you know if the two estimates are statistically different. You can impress your funders, by telling them whether the difference between the two data points is “statistically significant.”
Here are some examples of how you can write about statistical significance in a grant report:
“From 2010 to 2014 the poverty rate in area X was between X% and X% which is significantly higher than the poverty rate for Orleans Parish at the 90% confidence interval.”
“From 2010 to 2014, X% of residents had less than a 9th grade level, which is significantly higher than in 2000 at the 90% confidence interval.”
Source: American Community Survey (ACS) Guide. Metropolitan Area Planning Council. Retrieved May 17, 2012 from http://www.mapc.org/sites/default/files/MAPC-Guide-to-American-Community%20Survey.pdf
Note: For technical users, note that the Census 2000 Summary File 3 data has a small margin of error that is not provided in our neighborhood statistical area profiles. As a result, in rare cases, the widget may provide a false positive for statistical significance. For information on how to calculate the margin of error for 2000 data, please see Chapter 8 of the Census Bureau’s Technical Documentation available at http://www.census.gov/prod/cen2000/doc/sf3.pdf