data classification

Question 1

Natural Breaks

Answer 1

ArcGIS Pro default
• Looks for groupings in the data
– minimizes variation within each class – identifies ‘big jumps’ in the values
• Values within a class are likely to be similar
• Good for data that are not evenly distributed
• Difficult to compare different maps
• Difficult to choose appropriate number of classes

Question 2

Quantile

Answer 2

Equal number of observations in each class
• Good for evenly distributed data
• Emphasizes the relative position (e.g. which counties are in the top 20 percent)
• But: Features with similar values may end up in different classes, esp. if values cluster
• A few wide ranging adjacent values may end up in the same class

Question 3

Equal Interval

Answer 3

• Each class has an equal range of values
• The difference between the high and low value is the same for each class
• Easy to interpret
• Good for mapping continous data; no gaps or missing values in classification
• Good for comparison of a series of maps
• Inappropriate if values are clustered: there may be many features in one or two classes and some classes with no features

Question 4

Standard Deviation

Answer 4

Each class is defined by its distance from the mean value of all features
• Good for displaying features above or below an average value
• Good when the distribution is normal
• But: Does not show actual values of features, only their distance from the mean
• Very high or low values can skew the mean so that most features will fall into the same class

Question 5

Unique (Manual)

Answer 5

• Class boundaries set in accordance with external criteria
• E.g. state requirements; specific criteria set by researcher
• Requires an understanding of the broader context of the data