Thematic map - population change in the U.S.

https://www.e-education.psu.edu/natureofgeoinfo/book/export/html/1553 

 

The theme depicted in this thematic map of the U.S. is population change across time.  In this case, the change is shown in percentages by county, over a 10 year span (1990-2000).  This thematic map can more specifically be referred to as a classed choropleth map – the shadings represent different ranges of percentage of change.  Thematic maps like these are very useful for examining a phenomenon and its occurrence across space.  We are able to see that during this time period, the population increased the most drastically in the West, particularly in Arizona and Washington State, as well as in Florida. 

Mental map - Boston

http://ocw.mit.edu/ans7870/MAS/MAS.965/f04/assignments/tripti/Assignment10/index.htm

 

This is quite obviously a mental map for several reasons.  First of all, it is hand-drawn.  Secondly, it has no fixed scale – buildings and streets are color-coded, but we are not sure of the distances and proportions.  This mental map of a part of Boston serves a certain purpose:  not necessarily an accurate depiction of the whole picture, but rather a workable picture of the necessary landmarks in order for a viewer to navigate him- or herself, as determined by the person who drew the map.  Mental maps, though subjective, can be very helpful in day-to-day activities, as well as in determining what things characterize a neighborhood, or stand out in memory.

Star plot - MER ID sample designs

http://start1.jpl.nasa.gov/caseStudies/autoTool.cfm

 

In this star plot, four different design samples (MER IDD) are compared and sorted by seven metrics variables (accuracy, collision, trajectory completion, time, mass, actuator saturation, and link deflection).  According to the description underneath, in this star plot, the center represents the most desirable result. Each design represented on the plot has a different color; in red, they also added a handcrafted MER ID design, for comparison and validation purposes.  It is clear that each design has its setbacks and its advantages, so the way to determine the best design is to decide which metrics are the most important, and then to see which design fulfills these metrics the best.  In this case, the website description says design samples 3 and 4 fulfill the important metrics the best.

Correlation matrix - Cancer types

http://www.livestockgenomics.csiro.au/Genome_to_Phenome/  - click on Figure 8

 

This correlation matrix is examining possible correlations between types of cancers.  In a typical correlation matrix, there is a diagonal line of correlation=1 (in this case red, meaning that brain cancer correlates with brain cancer, etc.)  What we are looking for are other red boxes underneath and above the diagonal line.  It is interesting and perplexing to see that “Brain_Normal” is seemlingly correlated with “Brain_Cancer”, “Breast_Cancer” and “Colon_Cancer”.  However, since this diagonal correlation line is beneath the standard red correlation line, it means these variables are negatively correlated – so, if you have a “normal” brain, you are likely not to have brain cancer, breast cancer, or colon cancer.  Again, it is important to note that this matrix depicts correlation, not necessarily causation.

Similarity matrix - behavioral malware clusters

http://cgi.mtc.sri.com/Cluster-Lab/

 

In this similarity matrix, we can see a diagonal line of red boxes, which decrease in size from left to right.  These red boxes represent a strong similarity (close to identical).  The diagonal line of red boxes is surrounded by mostly blue boxes – this color indicates small similarity.  These similarity matrices can be used for various purposes; this one is being used to determine malware behavioral patterns, and hopefully to predict future patterns by their similarities.

Stem-and-leaf plot - height in elderly women

http://what-when-how.com/statistics/skewness-to-systematic-review-statistics/ 

 

This stem-and-leaf plot shows us that of the 351 elderly women in the sample, most of them are either 158 or 159 or 163 inches tall.  At first, I had trouble understanding the system, but once I did, it was very easy to interpret the results:  the “stem” portion on the left stands for the first 2 digits of the height, while the “leaves” represent the heights of each woman in the sample.  If you turn this stem-and-leaf plot around to be vertical instead of horizontal, you get a histogram.

Box plots - foreign country visits

http://cnx.org/content/m46174/latest/?collection=col11521/latest

 

Box plots are helpful for depicting groups of numerical data through their quartiles.  This box plot compares three countries – Germany, America and China.  Specifically, the data shows results of a survey of 20 year-olds in each country when asked how many foreign countries they had visited in their lifetime.  We can see that 20 year-old Germans have visited the most foreign countries (between 4 and 8 countries), with American 20 year olds behind (between 0 and 5 countries) and Chinese 20 year olds without any data (which I assume means 0 countries).