Metadata for Datasets with filetype=kni: Key Neighborhood Indicators

Rev. June 20, 2017

See the Readme.html file in this directory for more background information.

Click on a variable name to scroll to its description:

ur2 | state | county | fipco | tract | esriid | sumlev | cbsa | cbsatype | cbsayr | csa | metdiv | intptlat | intptlon | landsqmi | totpopsf1 | over15 | over1 | over25 | tothhs | ua | uaname | puma | placename | pcturban | totpop | areasqmi | medianage | pctunder18 | pctover65 | pctblack2 | pctindian2 | pctasian2 | pctnonhispwhite | pcthispanicpop | medianhhinc | meanpovratio | avgfaminc | pctpoor | pctpovratio1to2 | pctunemployedclf | pctnotinlf | avgcommute | pctmanprofoccs | pctagriculture | pctmanufacturing | pctnotinsuredunder65 | pctfamhhs | pctfamswithkids | pctsinglemothers | avghhsize | pcthhswithelders | pctgrpquarters | pctdivorced | pctnevermarried | pctincollege | pcthsorless | pctbachelorsormore | pctdiffhouse | pctforeignborn | pctvachus | pctunits1 | pctmobilehomes | pctbuilt2010orlater | pctbuiltbefore1940 | pctmovedin2010orlater | medianhvalue | pctrenterocc | mediangrossrent | employmentestimate | rankuniv

Geographic Variables

Urban/Rural category. Value of '1U' indicates a tract that is at least 50% urban. Otherwise value is '2R' indicates a (mostly) rural tract. Based on portion of population in Urban/rural areas per the 2010 census. See also PctUrban.

FIPS state code.

Stored as a 5-character FIPS code with leading zeros, but has a format associated with it so that it displays as the name of the county. Reflects changes since 2010 in Alaska, SD, and VA.

FIPS county code - same stored value as County (previous) but does not get displayed as the name but just as the code.

2010 Census Tract (revised). This is the "neighborhood". Code stored in format with leading zeros and '00' suffixes. So a tract 12 would have a value of '0012.00'. "Revised" means that where tracts have been renamed or redrawn post 2010, we use the new code. This allows matching to the ACS data, but creates minor problems if matching to 2010 census data.

Geographic ID variable that combines the state, county and tract codes. Used as a key to facilitate easy matching with shape files provided by ESRI (Arcview/ArcInfo company). Tract 12 in state SS, county CCC would have a value of 'SSCCC001200' with tract as the actual tract code without the period.

Geographic Summary Level. Should always be 140.

Core Based Statistical Area - i.e. Metropolitan Statistical Area (MSA) or Micropolitan Statisital Area. A 5-digit code with leading zeros. A value of 99999 indicates outside any CBSA.

Values are 'Metro' , 'Micro' or blank to indicate a Metropolitan Statistical area vs. a Micropolitan SA. Blank indicates not in any metro/micro area (CBSA=99999).

Year of the CBSA definition. Currently has a value of 2015.

Consolidated Statistical Area. These entities are comprised of adjacent and related CBSAs. A value of 999 indicates that no CSA is defined.

Metropolitan Division. This is a sub-area of a CBSA. A value of 99999A indicates that no Metropolitan Division is defined here.

Internal point Latitude . A point located at the centroid of the tract (unless the centroid is outside the tract in which case it becomes the nearest point on the tract boundary). In decimal degrees with 6 decimal places displayed. This comes from the Census Bureau's TIGER database.

Internal point longitude. This is the "x" coordinate that goes with the latitude (intptlat) value. It is a negative value, indicating a west longitude.

Land area in square miles as provided by the Census Bureau. Excludes significant water areas.

2010 Census Total Population count. (OK, it's not a geographic item -- but it didn't fit in any of the other categories either.) The "sf1" suffix denotes the origin of this value being Summary File 1 from the 2010 census.

More (Secondary) Geographic Variables

Urbanized Area or Urban Cluster code. Unlike CBSAs, these do not change over the decade. Tracts can be only partially within a UA. The UA code and UA name items refer to the UA with which the tract has the greatest intersection. The PctUrban variable indicates what portion of the tract was in the UA as of the 2010 census. Blank when the tract is totally rural (the same as saying it does not intersect with any UA).

Name of the Urbanized Area. Goes with the UA code variable.

PUMA: Public Use Microsample Area code. This is the 2010 (aka "2012") PUMA as used to tabulate ACS data with vintage 2012. Tracts will never cross PUMA boundaries. Every tract is in one and only one PUMA. So they make good universes for displaying or analyzing tracts.

Tracts can cross place (municipality or Census Designated Place) boundaries. This variable contains the name of the place with which the tract had the greatest intersection at the time of the 2010 census. A blank value indicates the tract is entirely outside of any place. (The variable is actually a 7-character string containing the 2-character state code and the 5-character FIPS place code, and is associated with the $place format code causing it to display as the name.)

This is the percentage of the tract's population that was classified as being urban per the 2010 census population counts. See the UR2 variable which gets assigned based on the value of this measure. This value never gets updated over the decade.

Total Population - the 5-year ACS period estimate.

Area in Square Miles. Includes water areas (unlike the LandSQMI variable.)

Universe ("Weight") Variables

These variables are not key indicators but are included for the sake of being used to weight certain percentage variables when aggregating or to derive a count from a percentage. For example, Over25 is the universe variable for the educational attainment indicators (all percentages). You can detrmine the count of persons with Bachelors degrees or more using the formula: CountBachelorsorMore=PctBachelorsorMore*.01*Over25 .

Persons 15 years and over. Universe for marital status variables.

Population 1 year and over. Universe for one-year-ago items.

Persons 25 years and over. Universe for educational attainment variables.

Total households. Universe variable for lots of household-based indicators.

Key Indicators

Key Indicators Come in Sets of Five

Each of the key indicator variables described here come with the base value plus 4 statistical measures of how these values rank compared to other tracts in the ranking universe. The variables containing these measures have names formed by appending the suffixes to the base variable name (so that PctPoor_pcat is the variable with the percentile category value for PctPoor). These measures are as follows:

  1. _pcat: The "percentile category" code for the variable. These codes are the essence of identifying key indictors. The possible values are:

  2. _pctl: This is the actual percentile rank of the indicator. We used average values in case of ties.

  3. _Z: This is the "Z score" for the indicator. If you are savvy regarding statistics then you will know what this means, and if you are not then you can ignore these values. (It is the number of standard deviations from the mean, with negative values indicating below the mean).

  4. _mdiff: This is the difference between the mean value for the indicator and the mean (average) value for all tracts in the ranking universe. So if PctBachelorsorMore_mdiff has a value of 3.8 and PctBachelorsorMore is 32.4 it says that the 32.4 value is 3.8 percentage points above the value for the average tract (not exactly the universe average, but close). So x + 3.8 = 32.4, where x is the universe mean. Solve for x and you get 28.6 as the universe average.

Median Age. The median is the middle value. If a tract has 4001 people you would use the age of the 2001st oldest person as the median.

% persons Under 18 (children)

% persons over 65 (seniors)

% Black or African-American Alone or in Combination

% American Indian, Eskimo or Aleut Alone or in Combination

% Asian Alone or in Combination

% NonHispanic White alone . Race and Hispanic Origin are separate categories (Hispanic is NOT a race).

% Hispanic (of any race)

Median Household Income. The middle value of the household income (i.e. of all persons in the household.)

New for 2015! . This is the mean poverty ratio. Each person for whom poverty status is assigned lives in a household which has a "poverty threshold" associated with it. The threshold for a person living alone is much lower than the threshold for a family of 4 with 2 children. We think this is the best economic indicator available. We have gone to considerable trouble to create this variable based on data in the C17002 summary table combined with data from the ACS PUMS. We need to do a separate document to explain all this. Coming soon.

avgfaminc (deleted for 2015)
Average Family Income . Ignores the income of persons who are not living in family households.

% Persons Below Poverty .

pctpovratio1to2 (deleted for 2015)
% Persons With Poverty Ratio 1 to 2 . These are people above the poverty line but with income less than twice the poverty threshold for their household. Sometimes referred to informally as the "working poor".

% Unemployed (persons in the civilian labor force)

pctnotinlf (deleted for 2015)
% Persons Not in the Labor Force

Average Commute Time for persons commuting to work.

% Persons in Managerial Professional Occupations

% Persons in Agricultural Industry

% Persons in Manufacturing Industry

% Persons Uninsured Under 65

% Family Households (deleted for 2015)

% Families with Own Children

% Single Mother Households

Average Persons per household (deleted for 2015)

% Households w Elders (person(s) aged 65 or more) (deleted for 2015)

% In Group Quarters (nursing homes, dormitories, barracks, prisons, etc.)

% Divorced and not currently married. (deleted for 2015)

% Never Married (deleted for 2015)

% (persons over 3) Enrolled in college

% Persons Over 25 with High School education or less

% Persons over 25 with Bachelors Degree or more

% Persons in different house last year

% Foreign Born

% Vacant Units

% Single Family Units

% Mobile Homes

% Housing Units Built 2010 or later

% Housing Units built pre 1940

% Occupied Housing Units Moved into 2010 or later

Median Value Owner Units. Middle ("typical") price of a home.

% Renter occupied units

Median Gross Rent. Middle ("typical") gross (including utilities) rent paid.

Estimated Persons who worked in the tract in 2014 (allocated from ZIP business patterns data).

Rank Universe. This 2- or 4-character code indicates the universe of tracts used to determine the "pcat" values. The first 2 characters contain a 2-character state postal abbreviation if only data from the same state as the tract were used. It might also contain "us" to indicate that tracts in all states (and DC) were used. If the code is just 2 characters then we looked at all tracts in the geographic universe. If the value ends with "1U" or "2R" it means we only looked at other tracts with the same value of ur2 (basically at other tracts with the same Urban/Rural designation). Datasets come in pairs having the same geographic and time universes, with one having pcat value based on all tracts for the area/time and the other having pcat values based on the same area/time and urban/rural category. For example, we have data sets mokni15 (data for Missouri tracts vintage 2015 with pcat values based on all tracts in the state) and mourkni15 (data for the same Missouri tracts vintage 2015 but with pcat values based on ranking the tracts within urban/rural categories.

Missouri Census Data Center

MCDC file: /pub/data/kni/Metadata.html
Metahtml rev date: 6/20/2017 10:43:00 AM