Part I – pick a dataset on Many Eyes that already has at least one visualization and create a different visualization of that same data. Is your version better? Worse? Why? Write a short explanation (about one paragraph)
Data: Crimes in Lawrence, Kansas (1999-2006) / Data visualization: Bubble Chart
My choice: Stack graph.
I think that my version of the visualized data isn’t better, but different. I like that mine gives a better sense of crime rates over the years thanks to how stack graphs show a sum, whereas the bubble chart doesn’t do that with just the scale legend on the side. However, it doesn’t make full use of the given data as specific types of crimes are too recorded. What would significantly improve my data visualization is if I was able to categorize all the types of crime into its related levels of punishment for said offences under Lawrence, Kansas’ jurisdiction. For example, I would like my graph to group all possessions of drugs, trespassing, simple assaults and so on as misdemeanours. Felonies, federal crimes, infractions, and etc would also be mapped with the appropriate crimes like how
shows “human resources”, “net interest”, and “national defense”. This would be a nice addition of information as it gives a sense of the severity of the crimes that’s been committed. Come to think of it, maybe the scatterplot would have resulted in the best represent of the given data…
The simple stack graph worked better than the stack graph with categories in my case! I like that you can also toggle between the “total” and “percentage” for different perceptions.
Part II – create your own dataset by importing some data into Many Eye. Create a visualization for that data and explain why you decided to use this particular visualization type over the others. Did you discover any pattern in the data?
First I made this one, but with the line graph (http://manyeyes.alphaworks.ibm.com/manyeyes/visualizations/us-emissions-metric-tons-per-perso). I felt it wasn’t effective in presenting such a heavy issue. The line graph had a lot of empty, negative space. So I went with the stack graph once again. I think the bar graph would of worked too, but again, the negative space in between the bars seem contradictory to the presented issue. From the visualization of my data, I can see that the US, by themselves, contribute half of the total amount that the entire North America emit as a whole. Further observations show that North America contributes one third of the world’s emissions. I may not have realized this as easily if I was just dealing with the numbers.
and also, because I felt my first visualization was too simple:
Surprise, surprise. The States and North America are the largest meat consumers in the world. It was a given to use the world map as my data provide stats for the entire world. I prefer the colour code over the bubbles to demonstrate the data. From the data visualization I can also see that there’s a strange intense drop in meat consumption in 1995. Not sure why, would need to research that a bit to find out… Also, Russia did not/barely consume any meat until 1992. Greenland did not consume a vase amount until 2002. Mongolia is the most consist in their meat consumption. Czech Republic did not form until 1993. Good information!