Hollinger – Week 2

Chapter 1:

Chapter one was a good introduction and foundation for concepts the book explains more in-depth later in Chapters 2, 3, and 4. Some of the important terms were: discrete (the feature’s actual locations can be pinpointed), continuous (the features blanket the entire area you are mapping and aren’t pinpointed to one location), and summarized by area, categories (groups of similar things), ranks (features in order from high to low – you only know where a feature falls in the order, don’t know how much higher or lower), counts/amounts (counts – the actual number of features on the map, amounts – any measurable quantity associated with a feature), and ratios (show you the relationship between two quantities). Discrete and continuous are considered types of features, while categories, ranks, counts/amounts, and ratios are considered types of attribute values. Furthermore, categories and ranks are not continuous values, they are a set number of values in the data layer, and counts, amounts, and ratios are continuous values, each feature could have a potentially unique value in the range (highest to lowest). Thus, each of these types of attribute values can be classified as a certain feature type. Another key part of chapter one was the difference between calculating and summarizing in your data tables. Calculating allows you to assign new values to features in your table and summarizing allows you to take the values for certain attributes to get statistics. An example of summarizing would be calculating the total average mean. This is important as it is the basis of how you get your data values to work with in the GIS.

These features seem to be the foundation of GIS and classifying data because if you know how to determine which of these terms your data falls under then you will know the best ways to represent it when it comes to mapping. Ideally, Mitchell explains that understanding each of these terms and correctly classifying your data will lead to maps that better represent and display the patterns in which you are trying to see.

Chapter 2:

Chapter two takes the terms from chapter one and dives a little deeper into the why, what, and how you should map. In terms of what you should map, Michell discusses 3 key things. Knowing where your locations are, being appropriate for the audience, and being appropriate for the issue. This means things like not providing too much or too little detail and paying close attention to adding reference features like roads your audience may know to give them context. The next portion of the chapter discusses how to get your data ready to map (assign it coordinates and category values) and then the mapping itself. It goes over two types of mapping (mapping a single type and mapping by category). When mapping a single type, Mitchell recommended using the same symbol to represent all features, but you can also show a subset of features with category values. When mapping by category a different symbol should be used to represent each category. Creating a separate map or subset for each category may make patterns easier to see as well. I think the most important thing I took away from this part of the chapter is that the way you chose to represent the features can alter the patterns. For example, you should use no more than 7 categories because patterns become harder to distinguish. Additionally, using too small or too big of an area relative to your features can obscure patterns as well. Mitchell concludes by talking about how symbol colors and size as well as reference features can change and affect the look of your map as well. Overall, I thought this discussion of aesthetics was important because looking at some of the “wrong” examples in the book, you could tell they didn’t display the data as well as the “right” examples.

Chapter 3:

Chapter three was a long chapter, but it was pretty straight forwards. I found that in many spots it often just elaborated on the concepts and terms we had learned in previous chapters. It went back over discrete, continuous, data summarized by area, counts/amounts, ratios, ranks, and classes. Then it goes on to talk about different schemes you can use to determine your distribution values. There are 4 types: Natural Breaks or Jenks (a natural grouping of data values – breaks where there is a large jump in values), Quantile (each class contains an equal number of features), Equal Interval (difference between high and low values is the same for every block), and Standard Deviation ( based on how much their value varies from the mean). Each of these distributions creates a very different map because certain data points fall differently into the categories depending on which distribution you use. Mitchell then continues to show how you can use a bar chart to visualize the distribution and determine which classification scheme is best. He then talks about outliers and how they can skew the data so you should make sure that they are not a mistake and then group them into their own category or in with the rest of the data. The chapter then moves into an in-depth discussion of ways to show quantities on a map. These are Graduated Symbols, Graduated colors, Charts, Contours, and 3D Perspective Views. The book thoroughly discusses all of these and their appropriate uses, advantages, and downfalls. I personally didn’t like the way any of the chart maps looked. I felt like they displayed too much information and the charts were so small it was hard to read. The chapter was finalized with a discussion of what patterns to look for in maps. These included highest, lowest, clusters, scattered, and even distribution.

Chapter 4:

Chapter 4 was a lot shorter than chapter 3, but it went a lot more in-depth. Particularly it focused on density mapping. Mitchell discusses two ways to map density. The first is by defined area. In this method, when using a dot map each dot represents a feature and is distributed randomly in the given area. These dots DO NOT represent the exact locations of a feature. You can also graph the density value for each area. In this case, using too large or too small areas can skew your graph making patterns hard to see. The other method of mapping uses density surfaces. This uses a raster layer as discussed in chapter one and each cell gets an individual value. This process is much more detailed but takes longer. The chapter then goes into depth on dot density maps. The most important takeaway I got from this section was that the more each dot represents the more spread out they will be, and dots should not be so big as to obscure patterns. After this discussion concludes, the chapter then moves into the specifics of creating a density surface. This discussion includes what cell size to use, how large the search radius should be, and 2 calculation methods. The simple method counts only features within the search radius of the cell, while the weighted method uses mathematical functions to give more importance to features toward the center of the cell. Ultimately, the weighted method results in a smoother surface that is easier to interpret. The chapter then moves into how you should display the data and brings back the distribution models discussed in chapter three (natural breaks, quantile, equal interval, and standard deviation). They are applied to density in a similar way as discussed in chapter 3. The chapter discusses contour lines and how adding them to your density surface can provide clear labels and show variance across a region. This helps make patterns and feature clearer to the audience. The chapter finalizes on an important note that you should map the features on which you based the density with the density surface or on a separate map. I believe this is important in that it provides context to the viewer.