Thematic Mapping Principles

1
Thematic MappingPrinciples
2
Process

• What is the Purpose
• Exploration of data.
• Presentation of information.
• Determine what variables will be used.
• Examine the distribution of data.
• Determine if distribution or density is to be
  shown.
• Normalize be appropriate variables.
• Assemble map with 5 basic elements and supporting
  statistical data elements.

3
Data

• Values should be derived (ratios, rates or
  categories).
• Totals have been used be are generally not
  acceptable.
• Varying size of area alter the impression of
  distribution.
• Uniform distributions may be hidden.
• Values should be normalized by area.
• Normalized by area shows density while derived
  shows distribution.
• Know your data and its context.

4
Normalization

• Should be done only once, otherwise produces
  erroneous rates, ratios or categories.
• Can be done
• On-the-fly in most GIS by selecting two
  existing fields.
• Calculated and stored in a field in the table.
• Done by 3 methods
• By a total, which depicts the spatial
  distribution.
• By another value, which depicts the spatial
  distribution.
• By area, which depicts the spatial density.

5
Normalization

• Total
• Variable / Total Distribution
• Another Variable
• Variable / Variable Distribution
• Area
• Variable / Area Density
• Ambiguous
• (Variable / Total) / Area Invalid
• Have to do a multi-variable map.

6
Normalization
7
Selecting a Method of Classification

• Separates values into separate bins (categories).
• Four standard methods of classification
• Natural Breaks (Jenks)
• Quantile
• Equal Interval
• Standard Deviation
• Manual selection of bin size.
• Number of bins should be no more than 6.

8
Classifying Data
Opening the layer properties provides many
options for displaying data.
Color Options
Thematic Styles
Bin Ranges
There are many options for making several styles
of thematic maps.
9
Adjusting Classification Bins
Measures
Descriptive Stats
Distribution
Moveable Breaks
10
Natural Breaks

• Used for highlighting clusters or patterns.
• Done using the Jenks Optimization method.
• Analogous to a One-Way Analysis of Variance.
• Advantages
• Calculates natural groups of similar values.
• Assigns clusters in same bin
• Disadvantages
• Can be inconsistent in break calculations.
• Can hide peaks and valleys.
• Makes comparisons difficult

11
Natural Breaks
Data have extreme changes that are accounted for
using the Jenks method.
12
Natural Breaks
Hard to detect patterns are calculated use the
Jenks method.
13
Quantile

• Used when values are evenly distributed.
• Arranges values from lowest to highest and
  assigns an equal amount of values to each bin.
• Advantages
• Assures there is an approximate equal number of
  values in each bin
• Shows relative position to other values
  (percentiles.)
• Disadvantages
• Does not yield frequencies in bins.
• Differences between bins may be exaggerated.

14
Quantile
Bin are noticeably of varying sizes.
15
Quantile
16
Equal Interval

• Used for focusing on observation clusters around
  the mean in a normal distribution.
• Divides values into equal ranges of bins.
• Advantages
• Is easier to interpret.
• Good for continuous data.
• Disadvantages
• Non-normal distributions will cause many values
  to be assigned to lower/higher value bins.

17
Equal Interval
Focus is on those values around the mean (similar
values in space).
18
Equal Interval
Broken into equal ranges for comparison to
another variable.
19
Equal Interval
20
Standard Deviation

• Used for focusing on extreme values when the
  distribution is non-normal.
• Bins are 3 standard deviations above and below
  the mean.
• Advantages
• Shows variation above and below the mean.
• Emphasize outliers or clusters.
• Disadvantages
• Does not show actual values.

21
Standard Deviation
-1
2
3
gt3
-2
-3
lt3
22
Manual

• Used for focusing on specific criteria.
• Can isolate outliers or other anomalies, e.g
  bimodal distributions.
• Advantages
• Allows for specific bins to be created based on
  nature of research.
• Disadvantages
• Harder to interpret (basis for comparison).

23
Manual Selection
Set according to t-value significance levels.
24
Manual Selection
Isolates the extreme values in the bimodal
distribution .
25
Manual Selection
Separates the negative and positive t-values.
Set according to t-value significance levels.
26
Exercise

• Load data that was selected in earlier exercise..
• Explore the data embedded into the census
  geography.
• Make several density maps the percentage
  variables using the classification methods
• Any raw variable / AREA
• POP00_SQMI
• Make several distribution maps the percentage
  variables using the classification methods
• AV_HH_SZ Average House-Hold Size
• AV_FAM_SZ Average Family Size
• Can link data saved from Census application.