Small Area Statistics

1
Small Area Statistics
(Tiny, tricky geographies, and the people who
need them)
2
Geography
3
Geography and Statistics the Where

• Theres always a where ( time variables)
• The importance of the where varies depending on
  what youre examining
• Like statistics, where displays often spur new
  why questions

4
Why learn about small area stats?

• Small area statistics
• are essential for certain types of analyses
• can be challenging to find, understand and to
  work with
• can answer local and very specific questions
• can be expensive to produce and obtain (i.e.
  present access challenges)

5
Small Area Statistics

• Objectives today - to build understanding of
• The relationship between geography and statistics
• The terminology and hierarchical structure of
  Census geography, to understand commonly
  accessible smaller units
• Other small geographical units important to
  statistical display which are important/frequently
  requested
• How to use key Statistics Canada tools to find or
  generate spatial display of statistics

6
What are small area statistics about?

• High demand for information at the lowest
  geographic level available
• Statistics at sub-provincial, or sub-municipality
  level, are critical to analyses of
• health (e.g. spread of disease),
• housing, crime, social issues (e.g. emerging
  patterns of concern or interest),
• emergency preparedness (analysis of this doesnt
  work at a whole-municipality level),
• market analysis, (why do they want my postal code
  anyway?)
• and much, much more!

7
Privacy and confidentiality

• Keeping the unit of analysis anonymous is a
  challenge with small area information (if one has
  good local knowledge, you can identify a person)
• There are rules in place about what population
  counts are required in order for small area
  statistics to be released (e.g. income)

8
Expense and access

• Authoring agencies, because of budget
  limitations, (priorities) are always balancing
  availability of variable detail and finer levels
  of geography

9
Expense and Access

• More variables?
• Smaller geography? !!
• No access/distribution infrastructure in place
• Simply not available or collected

10
Geographic display

• What can a map display do that a listed table
  cannot?
• Summarize the big picture with a picture
• Rapidly show PATTERNS of disparity that might
  have some unexpected explanation
• Allow display of statistics without knowledge of
  coding structure for viewers

11
Brief words on why Geographic Information
Systems?

• Small area statistics are not easily read in
  tables
• Graphic display becomes much more important at
  smaller levels
• GIS increasingly used as a tool for small-area
  analysis and summary

12
John Snows Cholera Map
13
No need to label the areas the image says enough
14
Geographic displays of stats always involve
choices, too

• Simple shade of color choices imply different
  meanings
• Ranges of statistics (breaks in the data) can
  be manipulated to imply different things
• Statistics can be left out of maps easily what
  is missing?
• Source statistics may be bad

15
Sierra Club Deforestation
16
Geographic displays of statistics are subject to
metadata review

• Evaluation of an online map display is as
  required as an evaluation of statistics via
  metadata review metadata criteria also apply to
  maps (sources should be cited, survey specified
  see yesterdays slide)

17
The Census Geography Hierarchy

• Organizing a national system of statistical
  reporting depends on a full-coverage nested
  geographic hierarchy i.e. geography/GIS for
  StatsCan is about more than making maps
• The hierarchy helps to ensure 100 coverage of
  the population during Census collection by
  organizing the countrys geography
• The hierarchy also defines level of the release
  of statistics
• Small area statistics exist at the bottom (yet
  ) end of the hierarchy

18
The Statistics Canada Hierarchy
19
Great StatsCan Geography Tools for understanding
hierarchy

• Nice quick tutorial
• http//geodepot.statcan.ca/Diss/Reference/Tutorial
  /HC_tut1_e.cfm
• Fantastic glossary http//geodepot.statcan.ca/Dis
  s/Reference/COGG/Index_e.cfm

20
A brief hierarchy overview

• All levels of the hierarchy have definitions and
  corresponding codes
• Eg. Canada 00 Alberta (Province) 48
• The levels and codes have defined relationships
• Below provinces, we have Census Divisions eg
  4801
• Below provinces, Census Metropolitan Areas and
  Census Subdivisions
• Below those, Census tracts and Dissemination
  Areas (SMALL AREA STATISTICS)

21
Hierarchy continued

• Hierarchy is defined administratively (ie
  political decision) and statistically (ie
  StatsCans reporting requirements)
• Not everything in the hierarchy relates to every
  other unit (see chart) i.e. not a straight,
  linear hierarchy
• Eg. Forward Sortation Areas
• Odd units (to StatsCan) Designated Places

22
Hierarchy applied to statistics

• Not all statistics are available for all levels
  of the hierarchy parts of the hierarchy may not
  exist in some places
• Statistical analysis is more appropriately
  applied to some units than to others eg. CMA vs
  CSD

23
Hierarchy and Small Area Statistics

• What are the important small area statistics in
  the hierarchy?
• Most commonly Census Tracts and Dissemination
  Areas

24
Other small-area data units

• What people want re geography is often not the
  unit of geographic availability
• Data typically compiled into statistics to meet
  the needs of the authoring organization
• Who ELSE cares about these areas/what demands are
  in place for this information?
• Solutions are available! (Well look at some)

25
Neighbourhoods

• Frequent need for statistics at this level of
  geography
• Census tracts vs. neighbourhoods
• Municipalities purchasing profiles and sharing
  agreements

26
Postal Codes

• Frequently requested for market analysis/business
  applications
• Represented graphically by dots in a product
  called the Postal Code Conversion File (PCCF)
• Postal codes are regions!
• The PCCF allows matching of postal codes to the
  best corresponding dissemination area

27
Roads and their attributes
28
Unavailable Statistics Canada geographic areas

• Some data resellers impute or calculate
  estimates of missing data values for small area
  statistics

29
GeoSuite Walk-Through Exercise

• Click-along with Leah!

30
Exercise Explore the hierarchy and statistics
for your favorite geographic area using StatsCan
Tools

• Start here
• http//geodepot.statcan.ca/Diss/Maps/Maps_e.cfm
• Explore the three sources available and evaluate
  them for usability, metadata, and for what
  information they have to offer you
• What were you able to discover about your chosen
  area from each source?
• To what level of geography were you able to reach
  using each source?