ACS Application Usage Notes

(Novice users: Please view the MCDC ACS applications tutorial for an introduction to the Profiles and Trends menu and reports.)

Using the Profiles Menu

Use the profiles menu to select up to four geographic areas for a single ACS time period (year or years) and display a profile report for the selected area(s) and period.

Step 1: Select Period

ACS data are released three times each year: one set containing single-year estimates, one set of 3-year estimates, and one set of 5-year estimates. For example, in 2012 the Bureau released ACS datasets for 2011 (1-year), 2009-2011 (3-year), and 2007-2011 (5-year). The different period length datasets have different geographic coverages and different statistical reliability. (See the statistical reliability section below for more information.) Choose a period that's appropriate for your geographic and statistical needs.

Note that you can choose only one period to display on the profile report. If you change your period choice, all previously selected areas will be cleared.

Step 2: Select Area Type

Areas include all locations and types (states, counties, etc.) available for a given data period. These area types correspond to the Bureau's summary levels. An area type's availability will depend on the period selected, as noted above. For example, census tracts are available only for 5-year ACS data. For more information about geography types, refer to the Bureau's ACS geography reference page.

Step 3: Select State or Geographic Component

If you selected a large area type (generally state or larger) in the previous step, you'll now see an option to include geographic components. A geographic component is the portion of an area that meets a location-based test, such as "in a metropolitan area" or "in a rural place". In most cases, the default (to not show components) is most useful.

If you chose a smaller area type, Step 3 will allow you to filter selectable areas by state. This is a convenience for areas like counties, where you won't want to scroll through all the counties in the U.S. If you would prefer not to filter areas by state, select "all states" (at the top of the state filter select list).

Step 4. Select Areas

Click any geographic area on this list to include it in the profile report. Selected areas will appear on a list to the right. De-select an area by clicking the small X button next to its name in the selected list.

You may select up to four areas, in all, of any type. For example, you can select the whole U.S. from area type Nation, then Missouri and Kansas from area type State, then Kansas City from area type Metropolitan Areas.

Step 5: Show Subjects

The profile reports include four subject groups by default: demographic, economic, social, and housing. You may choose to exclude any subject group. Note that if you exclude a subject group from the profile report, it will also be excluded from the Excel and PDF versions of the report.

Reading a Profile Report

Profile reports include a heading that identifies the period of the report (e.g., "ACS Profile Report: 2011 (1-year estimates)") and the selected geographic area(s) being summarized. Geographic areas include both area name and the standard geographic code for the area (geoid).

The geoid is the same code used by the Census Bureau in their datasets. The general format of a geoid is:

  1. The three-digit geography type (summary level) code
  2. The two-digit geographic component code (usually "00")
  3. The letters "US"
  4. The area's FIPS code (usually comprising the FIPS state code + FIPS place code)

For example, the geoid code for Boone County, Missouri is 05000US29019:

  1. Summary level: 050 (county)
  2. Geographic component: 00 (not a geographic component)
  3. US
  4. FIPS state code: 29 (Missouri)
  5. FIPS place code: 019 (Boone)

Table Layout

The report contains a series of up to four subject groups (demographic, economic, social, and/or housing). For each individual subject table, the gray header area shows the subject title, the table universe (i.e., the basis for the counts and percentages in the table), and links to the ACS base table(s) used to generate the profile table. These links point directly to the U.S. Census Bureau's American FactFinder application, and may be used to obtain more detailed information.

Variable names in the leftmost column are indented to show data heirarchy. For example, in table D2. Age and sex, the label for "Male" is indented relative to "Total population". The percentage shown for males is the percent of the total population that are males. Similarly, the next label, "18 years old and over", is indented relative to "Male"; its number is the count of all males that are 18 and older, and its percentage value is the percent of males (not total population) that are 18 years and older.

Viewing Statistical Reliability of Numbers

The profile report shows statistical reliability of numbers at a glance by using heavier or lighter fonts to indicate values which are notably reliable or unreliable. The ACS base tables include a margin of error (MOE) value for each variable. We use these MOE values to calculate a relative MOE (RMOE) value, defined as the MOE as a percent of the numeric estimate. The profile reports indicate RMOEs as follows:

  • Bold font: RMOE less than 15% (most reliable)
  • Normal font: RMOE between 15% and 35%
  • Light font: RMOE 35% or greater (least reliable)

The corresponding percent column value is displayed with the same font/color as the number value.

To see the actual RMOE and confidence interval for a number, hover the mouse pointer over it to show a pop-up tooltip. The bounds of the confidence interval are simply the value plus and minus the MOE.

Other Formats and Metadata

At the upper right of the report there is a box containing links to related pages.

  • Export data to Excel will create an Excel file with each (selected) subject group on a separate worksheet.
  • PDF version will create a bookmarked PDF file containing all selected subjects and data.
  • Variable metadata links to a listing of all the variables in a profile, along with the formulas used to derive them. A profile variable or table may use source data from several different ACS base tables; this metadata listing shows which ones.
  • Extract data via Dexter links to MCDC's Dexter web application for accessing our data collection. This link takes you directly to the data extraction utility's front end with the appropriate dataset already selected.

Data Accuracy and Statistical Reliability

ACS data are survey estimates created using sample survey instruments, not enumerations created by counting the whole population. Statistics from all sample surveys are subject to sampling and nonsampling error. This is why each ACS data point includes a corresponding MOE. Changes in survey design from year to year can also affect results.

The ACS releases datasets three times each year: one set containing single-year estimates, one set of 3-year estimates, and one set of 5-year estimates. For example, in 2012 the Bureau released ACS datasets for 2011 (1-year period), 2009-2011 (3-year period), and 2007-2011 (5-year period). The different period length datasets have different geographic coverages and different statistical reliability:

  • The 1-year period estimates are more current, but are less statistically reliable and cover fewer geographic areas — only areas with populations of 65,000 or more.
  • The 5-year period estimates are the least current, but have the best statistical reliability and cover all geographic areas down to the block group level.
  • The 3-year period estimates offer moderate currency and reliability, and cover all areas larger than 20,000 population.

Choose a period that's appropriate for your geographic and statistical needs.

The Bureau prefers to use the Coefficient of Variation (or "CV") statistic as the relative measure of statistical reliability. The CV is defined as the SE (standard error) as a percentage of the estimate. The SE can be easily derived from the 90% level MOE using the formula SE = MOE / 1.645. So, the difference between the RMOE that we use and the CV is simply CV = (MOE / 1.645) / E = RMOE / 1.645, where E is the estimate and RMOE is the relative margin of error. The intervals we use to determine our three-tiered font scheme can be expressed in CV terms as:

  • Bold font: CV less than 9.1 (15 / 1.645)
  • Normal font: CV between 9.1 and 21.2 (35 / 1.645)
  • Light font: CV 21.2 or greater

The Census Bureau's ACS guidance for data users page has much more information about the different ACS products.