Class 3- GIS, Maps, and Spatial Analysis

Question 1

Thematic maps

Answer 1

• Goal: A thematic map is a type of map especially designed to show a particular theme connected with a specific geographic area
• communicate the distribution of one or more properties
  
  • Usually shown at small scale
    
    • Small scale equals large areal extent
  • Not restricted to any subject matter

Question 2

Abstraction

Answer 2

• Abstraction is the process of simplying/abstracting the earth so that it can be mapped
  
  • Levels of abstraction for information
    
    • Real world → Data Model → Data Structure
  • The Earth is infinitely complex and also infinitely large
    
    • Far too much information to model the world perfectly (extent and detail)

Question 3

Spatial data

Answer 3

• Spatial data is special
  
  • Linked geometric and tabular data
    
    • Position and attributes
      
      • Geometric data are different than adding a “location” name or coordinate values (link)
  • Ability to integrate spatial and statistical analysis
    
    • Spatial: Distance, Direction, Adjacency, Containment
      
      • Advanced GIS and spatial analysis functions are built upon these primitive relationships

Question 4

GIS

Answer 4

• Geographic Information System
  
  • The systems, science, and technology associated with processing spatially referenced data to provide meaningful information to decision makers and other users (e.g., scientists)
  • A technology with links to other geospatial technologies: global positioning systems (GPS), remote sensing, information visualization
  • Defined by the functions it provides: data input, data storage and retrieval, data manipulation and analysis, and reporting/communication

Question 5

Geographic coordinates

Answer 5

• On a sphere, referencing is made using angles, using the center of the Earth as the origin
  
  • Latitude
    
    • Measured N-S relative to Equator
  • Longitude
    
    • Measured E-W relative to Prime Meridian

Question 6

Vector data

Answer 6

• Point, line, and polygon features represent entities
  
  • Many attributes can be stored for each feature
  • Features have to be defined and mapped
• Feature classes:
  • Points, Lines, Polygons
    
    • Points are the basic building blocks of vector data
  • Generally, each vector data layer contains a single “feature class”
    
    • Data layers contain multiple features (or discrete objects)
    • Each object may have multiple attributes

Question 7

Foundation data

Answer 7

• Foundation data is the “background” spatial data used as reference
  
  • Also referred to as “base map” data
  • Creating and/or developing spatial data requires a large amount of time, effort, and money
    
    • Good news: a large amount of foundation data has already been created and is freely available on the Internet
    • Bad news: the availability and quality of this data can vary significantly from place to place

Question 8

Geocoding

Answer 8

• Define location in “geographic space” using an address value
  
  • Also called address matching
• Requires a street/road database with specific attributes
  
  • Street name and number
  • Zone (in the US, a ZIP code)
• Attributes of the “event” are compared to the possible values in the street database
  
  • When/if a match is found, the event is assigned coordinate values in geographic space
    
    • Converts data from a list of addresses to point locations that can be mapped and analyzed
    • Very powerful tool in public health studies

Question 9

Table join

Answer 9

• Often, available health data is in a non-spatial data format
  
  • Tables!
    
    • Tabular data may be joined or linked to spatial features (foundation data)
    • Requires a common field
    • Requires a “unique identifier” for each feature
      
      • e.g., Name, ID, code
      • Not always easy… unfortunately

Question 10

HIPAA – IRB

Answer 10

• Health Insurance Portability and Accountability Act (HIPAA)
  
  • Health information that does not identify an individual and with respect to which there is no reasonable basis to believe that the information can be used to identify an individual is not individually identifiable health information.
• IRB
  
  • Institutional Review Board
    
    • Determines if studies/experiments are ethical

Question 11

Aggregation

Answer 11

• Individual data is aggregated to observation units
  
  • e.g., Census tracts, counties, states
• Why aggregate?
  
  • Privacy
  • Small numbers
  • Simplicity

Question 12

Ecological fallacy

Answer 12

• Ecological Fallacy occurs when findings from one “scale” are assumed to be true for another scale
• Ecological “scale” of analysis
  
  • Observation units are groups of people
  • Central tendency leveling
    
    • The more people in each observation unit, the less variation there will be among units
    • E.g., state, county, Zip Code, census block group

Question 13

MAUP – Zone and scale

Answer 13

• Modifiable Areal Unit Problem
  
  • Underlying data are aggregated to a single value (observation)
    
    • Masks variation within units
  • Location of boundaries is important
    
    • Size of the units and position
    • Two main effects from the MAUP
      
      • Zone and Scale
      • Affects statistical relationships

Question 14

Spatial pattern – Dispersed, random, clustered

Answer 14

• Clustered
  
  • Objects are configured or distributed near to one another
• Random
  
  • Objects are configured or distributed such that there is no regular pattern
• Ordered
  
  • Objects are configured or distributed in a regular or repeating fashion

Question 15

Spatial autocorrelation

Answer 15

• Spatial Autocorrelation
  
  • The degree of similarity between objects that are located near each other
  • Can be measured, quantitatively
    
    • Over an entire region (global)
    • In a smaller area within the region (local)
    • Use in health geography
      
      • Prediction, distance decay, cluster analysis

Question 16

Measuring Distance – Euclidean

Answer 16

• Euclidean distance
  
  • Straight line distance, based on Pythagorean Theorem
    
    • Assumes planar geometry – No changes in elevation
      
      • No impediments

Question 17

Distance decay

Answer 17

• Interaction or relationship among phenomena decreases with increasing distance
  
  • Distance has “friction”

Question 18

Reference maps

Answer 18

• A reference map is a map that emphasizes the geographic location of features. For these maps, the goal is to display a lot of different kinds of information without drawing the reader’s attention to any one theme of information more than any other theme.
• Goal: communicate spatial association between various phenomena
  
  • e.g., roads, settlements, administrative boundaries, water bodies, etc.
• Example:
  
  • USGS Topographic Map
  • Road map
  • World map of countries

Question 19

Dot density maps

Answer 19

• Shows density or intensity of some phenomena across space
  
  • Data are points / dots
    
    • Areal data can be converted to dots representation through symbology
• Dots are the same size

Question 20

Graduated symbol maps

Answer 20

• Shows intensity at specific locations
  
  • Size of symbol varies
• Also called “Proportional Symbol Map”

Question 21

Choropleth maps

Answer 21

• Show differences in intensity between areas
  
  • Health data is often only available at an “areal” level
  • Rate or ratio data
    
    • E.g., disease prevalence or incidence
  • Areas are shaded with different colors, patterns, or intensities
    
    • Sensitive to data classification scheme

Question 22

Projected coordinates

Answer 22

• Cartesian coordinates- also referred to as “projected coordinates”
  
  • Location referenced relative to two perpendicular axes (X,Y)
  • Measurements are in distance along each axis from the origin (0,0)
    
    • Most common units:
      
      • English: inches, feet
      • Metric: centimeters, meters, kilometers

Question 23

Raster data

Answer 23

• Field view: surface is continuous and every location as a value
  
  • One raster “layer” describes one attribute or characteristic
  • Attribute based
  • Uses grid cells
    
    • Regularly spaced grid cells (tessellations)
    • Each cell coded with a single value
• For some phenomena or attributes of interest, field-type representation seems more appropriate
  
  • When phenomena or attribute varies continuously across space
  • Cases when discrete representation does not adequately capture the nature of the phenomena
    
    • E.g., Weather or climate, land use / land cover, air pollution, elevation

Question 24

Ecological Fallacy Solution

Answer 24

• Test statistical relationships over multiple scales of analysis (levels of data aggregation)

Question 25

MAUP Zone Effect

Answer 25

• The zone effect is observed when the scale of analysis is fixed, but the shape of the aggregation units is changed.
  
  • Ex: analysis using data aggregated into one-mile grid cells will differ from analysis using one-mile hexagon cells.
• The zone effect is a problem because it is an analysis, at least in part, of the aggregation scheme rather than the data itself.

Question 26

MAUP Scale Effect

Answer 26

• The scale effect exhibits different results when the same analysis is applied to the same data, but changes the scale of the aggregation units.
  
  • Ex: analysis using data aggregated by county will differ from analysis using data aggregated by census tract.
• Aggregated units soak up some of the variability present in the original data hidden within the aggregated units

Question 27

Measuring Distance- Network (Travel) Distance

Answer 27

• Network distance restricts travel along existing arcs (lines)
  
  • e.g., on city streets
• Connectivity is the key concept
  
  • Space is undefined off of the network
• Distance is the sum of the lengths of arcs that make up the route

Question 28

Measuring Distance- Travel Time

Answer 28

• More accurate portrayal of human movement
• Requires travel network or travel surface
  
  • Also,software capable of this type of analysis (i.e., Geographic Information System)
  • Can be computationally intensive
• Vehicular-based : US studies
  
  • Augmented using alternative travel modes
    
    • e.g., public transit, walking

Question 29

Why Map?

Answer 29

• Nicer than looking at a data table
• Examine geographic variation
  
  • Both attributes and spatial relationship are visible
• Test for pattern
• Understand relationships, concurrence

Question 30

Tobler’s first law of Geography

Answer 30

• “Everything is related to everything else, but near things are more related than distant things”
• Values at locations near each other tend to be similar, with similarity decreasing with distance
  
  • Implies that phenomena are not distributed randomly (throughout space)

Question 31

LISA

Answer 31

• “Local Indicator of Spatial Association”
  
  • Local version of Moran’s I
    
    • Provides a measure of autocorrelation for each polygon
    • Returns “hot” and “cold” spots, as well as outliers
      
      • High-High, Low-Low, High outlier, Low outlier
  • Extremely useful for understanding “where” spatial autocorrelation is strong/weak

Question 32

Moran’s I

Answer 32

• Oft-used statistic for describing the spatial autocorrelation within a region
  
  • Global (considers the whole region)
  • Measures the magnitude of autocorrelation
    
    • Returns a single result (I)
      
      • Perfectly dispersed: -1
      • Random: 0
      • Perfectly clustered: 1