Assignment 1

Ian Godin
Geog 370
Assignment 1

Overview: This assignment serves as an introduction to statistical methods in geography. The first part of this assignment focuses on different classification methods and their uses. The second part of the assignment involves using these classification methods to represent data and make an accurate map

Goals: The goals of this assignment are to

• Differentiate between levels of measurement
• Differentiate between classification methods
• Retrieving data from the U.S. Census and joining data
• Enhance cartographic knowledge

_________________________________________________________________________________

Part 1:

The first part of this assignment includes defining different classification methods and providing examples of each. The four methods described below include Nominal, Ordinal, Ratio and Interval data. 

1. Nominal Data:

Nominal data is any type of data that does not have a number value associated with it. This may include city or street names, names of countries or symbols that represent location.

(Figure 1)

Nominal data map of the United States

The map in figure one published by the National Geographic Society is a great example of a map which displays nominal data. The map contains state names and uses single symbols to represent the locations of land forms such as lakes and rivers. No data with numbers associated with it are displayed. 

2. Ordinal Data

Ordinal data are types of numerical data that are displayed in some order. These data are typically displayed on a scale ranging from low to high. For example, a map displaying unemployment rates at the state level would have data arranged from low unemployment to high unemployment. Data can be either strongly or weakly ordered. Strongly ordered data is placed in a specific position within the order. Weakly ordered data are lumped together into categories and placed in an order.

(Figure 2) 

"Bikeability" map of the Vancouver Metro area

The map in figure 2, published as part of a study at the University of British Columbia, is a great example of a mag which uses ordinal data. The researchers collected data from bike riders and created an index to show how different factors make biking in the city more or less difficult. The data is presented on a scale of low to high with green representing areas that are easy to bike through and red representing difficult areas. 

3. Interval Data

Interval Data are types of data that are divided into equal intervals with no natural zero. This means that the data does not have a starting point. If the data does include a zero value, this serves as a point of reference. An example of this would be a map showing temperature data. This data can be divided into different intervals based on degrees and has no natural zero since temperature can go above or below zero degrees. 

(Figure 3)

Temperature map of North America

The map in figure 3, published by the Center for Sustainability and the Global Environment at the University of Wisconsin-Madison uses interval data. The map shows the average annual temperature across North America..The data is divided into intervals of 15 degrees Celsius. The zero value on the map is not the starting point but rather a middle reference point to show were temperatures fall along the interval scale. 

4. Ratio Data

Ratio data is similar to interval data, however this data has a natural zero point. Because of this it is possible to see the difference in ratios on the map. This classification is often used for population data as population is often represented as a percent or ratio of a whole. 

(Figure 4)

Map of Hispanic Population Density per Square Kilometer at the National Level

This map, published by Penn State University uses the ratio classification scheme. It shows the population density, or ratio of the Hispanic population at the state level.

Part 2:

The Second part of this assignment involves a fictional scenario where a company has asked for a map of the number U.S.D.A certified organic farms in Wisconsin at the county level. The goal of the map is to show which areas of the state would be best for promoting the development of additional farms. The challenge is selecting which data classification method would be best for displaying this information

Methods:

Using data from the 2012 census of agriculture and a shapefile of Wisconsin counties downloaded from the U.S census website I created three different maps each with a different classification method. Data was added to an Excel spreadsheet and joined to the shapefile using the Geo-i.d class. The data was then classified in ArcMap

Observations:

(Figure 5)

Equal intervals based on range

The first map uses the equal intervals based on range classification method. In this method, the smallest observation is subtracted from the largest and then divided by the number of classes to give the interval. In this case the largest observation was 233 while the smallest was one. This gives a range of 232 which when divided by four classes gives an interval of 58. This classification scheme is not ideal for this situation because the majority of the counties fall into the first class. This makes it appear as though all of the organic farms in Wisconsin are located in two western counties, which is not the case. 

 (Figure 6)

Quantile Classification

The quantile classification method breaks classes into intervals and places an equal amount of observations in each class. This map is better than the equal intervals method, however it may still be confusing to the reader. Due to the wide range of observation values, the class values have a very wide range. For example, a county in the fourth class could have as few as 20 farms or as many as 233. The data is not clearly represented. This map, however, does better show the distribution of farms throughout the state. 

(Figure 7)
Jinks Natural Breaks

In the Jinks natural breaks method, data is divided into classes based on naturally occurring gaps. This classification works better for this map because the naturally occurring gaps in the data allow for the observations to be nicely distributed into classes. This gives the reader a good idea of where farms are located throughout the state. 

Conclusions: 

I believe the best map to represent potential growth for organic farms would be the map using the natural breaks method. The values within the classes are nicely spaced apart and the reader is able to visualize the spatial patterns of where the farms are located. The best location to promote the development of organic farms would be in the central part of the state. This can be seen on the map where in the center of the state there is a large area of counties with small amounts of farms. This can also be observed in the far north and south-east parts of the state. These areas would not be ideal because the northern part of the state is mostly forest land (much of which is either privately owned or national forest land) and the south-east section of the state contains the many cities, notably the city of Milwaukee.