Lecture 8
Describe MAUP
A problem arising from the imposition of artificial units of spatial reporting on continuous geographic phenomena resulting in the generation of artificial spatial patterns
What is ecological fallacy?
A situation that can occur when a research or analyst makes an inference about an individual based on aggregated data for a group
What is a solution to MAUP
Be aware of how sensitive data is to MAUP
What does "basic" raster and vector analysis include?
A class of operations that are fundamental to a wide range of applications and are common in practice
What does MAUP involve?
Data that are spatially aggregated using arbitrary boundaries and affects our ability to quantify apparent relations, statistical or other
What type of data is rare to deal with in terms of spatially aggregated data
Data that is point specific
Give an example of MAUP
Electoral Districts and Voting
What does euclidean distance give and also leave out
Gives the distance but not direction
Where is gerrymandering common
In many countries, sometimes even legal (this is the case for the US)
How is electoral districts and voting an example of MAUP?
It's not just who you vote for, but also where you vote that counts - A different aggregation of US counties into states could have produced a different outcome (in the Al Gore/George Bush care)
Describe spatially aggregated data
Many geographic data are aggregates of data at a more detailed level
Discuss ecological fallacy in terms of stats
Median is a much stronger value than average, because usually the datasets are not normally distributed. They are normally skewed. Therefore, taking an average is not ideal because it would include the outliers
Describe dissolving
Merges points together to create an overlap
Describe Manhattan distance
More like a road network - moves around boundaries
Give some examples of spatially aggregated data
National census, traffic analysis zone, school district, watersheds, pixels
Even though we commonly classify spatial analysis as Vector or Raster...
Often the same analysis can be performed using either model, and many types of analysis involve both raster and vector data
How can we be aware of how sensitive data is to MAUP
Perform a sensitivity analysis to the MAUP by aggregating the data in different ways and seeing if the trends remain in all cases
Describe the Aggregation Effect
Take the number of units, and the same range of the units, and just change when the boundaries are
What are the 2 manifestations that MAUP includes
The Aggregation Effect and The Scale Effect
When dealing with buffering, what are we usually interested in?
The Euclidean (straight line) distance - How the crow flies
What does MAUP stand for
The Modifiable Areal Unit Problem
Describe the Scale Effect
The arbitrary-ness of the number of units you're using
What does vector buffering imply?
The calculation of distance
What is vector buffering?
The creation of a zone of interest around a spatial entity or set of entities
What is buffer width dependent on?
The individual characteristics of each feature
What specifically deals with ecological fallacy the most?
The media - Constantly draws conclusions about research done at one scale and applied it to a different scale
Describe gerrymandering
The practice of setting electoral district boundaries to favour a particular outcome
Why do we carry out analysis on geospatial data?
To uncover important spatial relations, gain understanding about some phenomenon or system, and perhaps to inform some decision
Describe MAUP when using spatially aggregated data
Usually the boundaries between one unit of aggregation and other are arbitrary
What are some buffering variables?
We can buffer points and lines
What is the main question you need to ask when dealing with dissolving
What are we particularly interested in? Do we want to identify some about the actual features/the zone/the whole aggregate data
When is ecological fallacy particularly a problem when dealing with data?
When dealing with data that is negatively skewed
When is ecological fallacy most significant?
When it is downscaled
When do we have to be especially careful for gerrymandering
When we modify the geographic boundaries of voting aggregation areas
What is one thing that is fairly consistent with MAUP?
When you aggregate data, you will have less variability than the underlying data - therefore toy will be able to determine trends in the data
What is one main rule concerning the ecological fallacy
You can't draw individual conclusions on large-scale findings - only at the scale of aggregation