1. ACS Public Use Microdata Samples DataFerrett SACOG
Luz M Castillo
Data Dissemination Specialist
Los Angeles Regional Office
U.S. Census Bureau

2. Outline Summary Data vs. Microdata Fundamentals of PUMS Data
Geography and the PUMS
Accessing PUMS Data
Documentation and Guidance

3. Summary Data Versus Microdata
Premade or published tables
Easy to get, even for small areas
Limitations: fixed content
Dataset of individual responses to questionnaire
Enables custom tables and analyses
Limitations: edits to protect privacy, can’t study small areas 3 3

4. Summary Data
Source: 2010 ACS 1-year Estimates. Table B FIRST ANCESTRY REPORTED 4

5. Microdata
Source: 2010 ACS 1-year PUMS file

6. Microdata in SAS
Source: 2010 ACS 1-year PUMS file.

7. Outline Summary data vs. Microdata Fundamentals of PUMS Data
Geography and the PUMS
Accessing PUMS Data
Documentation and Guidance

8. What are PUMS data? Public Use anonymized, downloadable Microdata
records of individual people
Sample
a representative sample of the population 8

9. PUMS Overview
PUMS sample is a subsample of ACS interviews, one percent of all US households
PUMS is a “weighted” sample
Weighting variables must be used in analysis
A set of two files - housing units and persons
Available as SAS files, CSV files, via DataFerrett and redistributors such as IPUMS 9 9

10. Why Use PUMS?
Data needed for a tabulation or a specific universe not supported by standard ACS tables (e.g., population groups by single year of age)
Statistical analysis required to understand relationships between economic, demographic or housing variables (e.g., correlation analysis)
Can create new measures using multiple variables or other people in household (spouse’s occupation, same-sex couples, number of kids) 10 10

11. ACS PUMS Availability Produced every year since 2000
Person-level files includes about 250 variables
Housing unit files include about 200 variables
Includes people in housing units and group quarters
Includes many useful constructed variables (e.g., poverty status, subfamily identification, etc.)
Includes collapsed codes for some variables (e.g., race, Hispanic origin, ancestry, place of birth, industry, occupation, etc.) 11

12. Person records in ACS PUMS
(millions)
Person records in ACS complete data (millions)
Population represented
2001
1.2
285
2002
287
2003
290
2004
293
2005
2.9
4.5
296
2006
3.0
298
2007
301
2008
304
2009
307
2010
3.1
309
2011
5.0
312 12

13. Types of PUMS Files Released
We release 3 new PUMS files every year
1 year PUMS (example: year PUMS)
October
3-year PUMS (example: year PUMS)
Discontinued after 2013
5-year PUMS (example: year PUMS)
January
Most documentation released one week prior to data 13

14. Modifications to Multiyear PUMS
Multiyear PUMS have the same cases and geography as their component 1-year files
How are multiyear PUMS different from single year?
Weights are produced using latest population estimate “vintages”
Coding schemes and dollar amounts are standardized
Why use the multiyear PUMS files?
For studying small groups, where more cases are needed
When analysis is also making use of multiyear summary data 14

15. Outline Summary data vs. Microdata Fundamentals of PUMS Data
Geography and the PUMS
Accessing PUMS Data
Documentation and Guidance

16. Limited Geographic Detail
Geographic identifiers are region, division, state, PUMA
PUMAs can be used to identify most cities of 100,000+ and many metropolitan areas, but not all
Combinations of adjacent counties and census tracts within states
Also, divisions of geo areas (counties/cities)
PUMS is not designed for statistical analysis of small geographic areas

17. Public Use Microdata Area (PUMA)
Defined after each census by the states in coordination with the Census Bureau’s Geography Division
Redefined PUMAs for 2012 PUMS files
Forthcoming multiyear files to have dual PUMA vintages
Large enough to meet disclosure avoidance requirements
An area of size 100,000 population or more
To determine population, housing, or land ratio visit the Missouri State Data Center site
PUMAs are identified by a five-digit number, unique within each state 17 17

18. Public Use Microdata Areas

19. PUMA Maps

20. PUMA Maps

21. 2010 Census – PUMA Reference Map: Sacramento City (Central/Downtown & Midtown)21 21 21

22. Outline Summary data vs. Microdata Fundamentals of PUMS Data
Geography and the PUMS
Accessing PUMS Data
Documentation and Guidance

23. American FactFinder 23

24. American FactFinder (cont’d)24

25. American FactFinder (cont’d)
Main benefit of accessing PUMS via AFF:
Convenient access if comfortable with AFF from regular use of summary tables

26. Census Bureau FTP Site

27. Census Bureau FTP Site (cont’d)
Main benefit of accessing PUMS via FTP:
Complete listing of files by year and state

28. DataFerrett 28

29. DataFerrett (cont’d) Main benefit of accessing PUMS via DF:
Menu driven system doesn’t require knowledge of a stats package (i.e. SAS, SPSS, etc.)
Ability to download variables individually 29

30. Powerful Tabulation Capabilities
Simple table layout that supports:
Flexible design
Frequencies and trends
Spreadsheet math for robust analysis
Complex nesting
Hide columns/rows
Applies weighting variables
Fast results using large datasets
Save as HTML, PDF & JPEG

31. Highlight spreadsheet rows or columns to create:
Data Visualization
Highlight spreadsheet rows or columns to create:
Maps
Graphs

32. What We’re Working On
Calculating variances on-the-fly for microdata tabulations
Calculating margins of error for custom summations of aggregate data
Integrating Google maps with DataFerrett thematic maps

33. Outline Summary data vs. Microdata Fundamentals of PUMS Data
Geography and the PUMS
Accessing PUMS Data
Documentation and Guidance

34. PUMS Documentation Subjects in the PUMS Code Lists
PUMS Top Coded and Bottom Coded Values
PUMS Estimates for User Verification
Accuracy of the PUMS 34

35. PUMS Guidance Compass Handbook on Using PUMS
soup-to-nuts overview of getting and using the data
Training PPT on Using PUMS
overview of PUMS basics

36. Exercise 1
In Placer County, how many foreign born individuals entered before 2000, between 2000 and 2009 and after 2010?

37. Exercise 1 – Nativity and Year of Entry
Access:
American Community Survey, Year Estimates PUMS
Foreign Born and Year of Entry Variables
Create a Recode for Year of Entry
All PUMAS within Placer County
Create a Table

38. Go to www.census.gov Type ‘DataFerrett’ in the Search Box

39. Click ‘TheDataWeb – DataFerrett’

40. Launch DataFerrett

41. CAUTION Do Not Navigate Away or Close This Window While DataFerret is Loading

42. Enter Your Email Address and Click ‘Ok’

43. Click ‘Get Data Now’

44. American Community Survey with PUMS and Other Datasets

45. Select American Community Survey Open Public Use Microdata Sample to view years Select Click View Variables (drop down)

46. Click ‘Selectable Geographies’ and ‘Population’
Click ‘Selectable Geographies’ and ‘Population’. Click ‘Search Variables’

47. Click on ‘Variable Label’ to Alphabetize Column

48. Select ‘Nativity’. Hold control button down and select ‘Year of Entry (YOEP)’. Click ‘Browse/Select Highlighted Variable’ (Blue Button).

49. Check the box next to ‘Select’ ACS Nativity’

50. Highlight ‘ACS YOEP’
Check the box next to ‘Select’ ACS YOEP Year of entry’
Click ‘OK’

51. You have added 2 variables for your DataBasket Click ‘OK’

52. Double Click to ‘Selectable Geographies’ Variable’ Click ‘Browse/Select Highlighted Variables’ (Blue Button)

53. Select ‘Public Use Microdata Area’ from ‘Types of Geographies Available’. Highlight the PUMA code in the Hierarchies section and click ‘Use Hierarchy’
Hierarchies

54. Double click ‘California’ from ‘Select State of current residence’
Double click ‘California’ from ‘Select State of current residence’. Highlight ‘California’ in middle box and click ‘Next Level’

55. Note: ALL PUMAs in California are Listed by County Double Click or Highlight and drag PUMA/s to box on far right. Click ‘Finish’

56. Note: There are 3 variables in DataBasket Click on ‘Step2: DataBasket/Download/Make A Table’

57. Highlight ‘Year of Entry’ variable
Highlight ‘Year of Entry’ variable. Click ‘Recode Variable’ from right side of screen

58. Rename ‘Recode1’ to ‘Year of Entry Recode’

59. Highlight the categories from ‘1921 to 1999’ and click ‘Recode’ button below

60. Highlight all of the categories from ‘2000 to 2009’ and click ‘Recode’ button below

61. Note: there are three categories for the new recoded variable
Note: there are three categories for the new recoded variable Change the ‘Label’ Names by double clicking inside the cells. (Make sure to hit the Enter Key when completed).

62. Note: ‘Year of Entry Recode’ now listed Click ‘Make a Table’

63. Click ‘OK’

64. You Will Now Make A Nested Table Using the Variables

65. Drag the ‘Geog-101 PUMA’ to ‘C1,R2’

66. Drag ‘RECODE1 Year of Entry’ to ‘C2,R1’

67. Nest ‘Nativity’ variable by dropping it onto any of the ‘Year of Entry Labels’.

68. Click ‘GO Get Data’

69. From File, Click ‘Save As’

70. You Can Save to Your Desktop Save File as Text Document – Comma Delimited (Excel)

71. Exercise 2
In Sacramento County, what age group under 50 has a higher estimate of individuals with a disability?

72. Exercise 2 - Age and Disability
Access:
American Community Survey, Year Estimates PUMS
Population with a Disability
Create a Recode for Age Disaggregation
All PUMAS within Sacramento County
Create a Pivot Table

73. Go to the ‘Step1’ Tab and Click ‘Empty DataBasket’

74. Select American Community Survey Open Public Use Microdata Sample to view years Select Click View Variables (drop down)

75. Click ‘Selectable Geographies’ and ‘Population’
Click ‘Selectable Geographies’ and ‘Population’. Click ‘Search Variables’

76. Click ‘Variable Label’ to Alphabetize Column

77. Select ‘Age’. Hold control button down and select and ‘Disability Recode’. Click ‘Browse/Select Highlighted Variables’ (Blue Button)

78. Check the box next to ‘Select’ ACS AGEP’

79. Highlight ‘ACS Disability Recode’
Check the box next to ‘Select’ ACS DIS Disability Recode’
Click ‘OK’

80. You have added 2 variables for your DataBasket Click ‘OK’

81. Note: 2 Variables selected in Data Basket Double Click to ‘Select Geographies’ Variable Click ‘Browse/Select Highlighted Variables’ (Blue Button)

82. Select ‘Public Use Microdata Area’ from ‘Types of Geographies Available’. Highlight the PUMA code in the Hierarchies section and click ‘Use Hierarchy’
Hierarchies

83. Double click ‘California’ from ‘Select State of current residence’
Double click ‘California’ from ‘Select State of current residence’. Highlight ‘California’ in middle box and click ‘Next Level’

84. Note: ALL PUMAs in California are Listed by County Double Click or Highlight and drag PUMA/s to box on far right. Click ‘Finish’

85. Note: There are 3 variables in DataBasket Click on ‘Step2: DataBasket/Download/Make A Table’

86. Highlight ‘Age’ variable
Highlight ‘Age’ variable. Click ‘Recode Variable’ from right side of screen

87. Rename ‘Recode1’ to ‘Age Recode’

88. Change Range to ‘1 through 17’ and Click ‘Recode’2
Change Range to ‘18 through 19’ and Click ‘Recode’ 3
Change Range to ‘20 through 24’ and Click ‘Recode’

89. Change the rest of the age groups and recode

90. Note: there are nine categories for the new recoded variable
Note: there are nine categories for the new recoded variable Change the ‘Label’ Names by double clicking inside the cell. (Make sure to hit the Enter Key when completed).

91. Note: ‘Age Recode’ now listed Click ‘Make a Table’

92. Click ‘OK’

93. You Will Now Make A Nested Table Using the Variables

94. Making a Pivot Table 1. Drag and Drop “Recode 1 Age” to C1, R2
2 . Drag and Drop “GEOG-101 to C2, R1
3. Drag and Drop “Disability” above R1

95. Click ‘GO Get Data’

96. From ‘File’ drop-down, Click ‘Save As’

97. Save Your Table

98. Exercise 3
For each race group, which age group has the highest estimated number of males and females?

99. Exercise 3 – Sex by Race and Age
Accessing:
American Community Survey, Year Estimates PUMS
Add more variables
Create a Race, Sex and Age Disaggregation
All PUMAS within Sacramento County
Create a Table, Chart and Map

100. Close Table and Click ‘Step1’ Tab

101. Select American Community Survey Open Public Use Microdata Sample to view years Select Click View Variables (drop down)

102. Click ‘Population’. Click ‘Search Variables’

103. Click on ‘Variable Label’ to Alphabetize Column

104. Select ‘RAC1P-Recoded Detailed Race Code’
Select ‘RAC1P-Recoded Detailed Race Code’. Hold control button down and select and ‘Sex’. Click ‘Browse/Select Highlighted Variables’ (Blue Button)

105. 1. Click ‘Select ALL Variables’
2. Click ‘OK’
3. Confirm that you have modified 2 Variables by Clicking ‘OK’

106. Click ‘Step2’ Tab and Click ‘Make a Table’

107. 1. Drag and Drop “RAC1P” to C1, R2
2 . Drag and Drop “GEOG-101 to C2, R1
3. Drag and Drop “Recode1 Age” to C1, R2 (On top of ‘Total RAC1P’)

108. Click ‘GO Get Data’

110. To Create a Bar Chart or Map, Change the Variable Label and Highlight the estimates in that row, Click the Chart or Map Icons

112. Resources: Need Assistance?
Data Dissemination Branch
Customer Liaison and Marketing Services Office
U.S. Census Bureau
(844) ASK-DATA Toll Free
Cell:
112
112