A pie chart representing spending by category for the US budget for 2007
Source: Wikipedia, 2007 United States federal budget.
link to the chart:
https://en.wikipedia.org/wiki/2007_United_States_federal_budget#/media/File:Fy2007spendingbycategory.png
Story Behind the Visualization
The above visualization was published in Wikipedia to provide information on the expenditures across different U.S government departments. The above visual is trying to convey the story of spending by the department for the US budget for 2007. There are a total of 31 U.S government departments and the pie chart is broken down with the respective department’s expenditures. The visualization has chosen percentage as the parameter to show the spending by the department for the US budget. The pie chart visualizations show that the highest spending is on the Social Security department, United States Department of Defence, Medicare, Medicaid and the State Children’s Health Insurance Program (SCHIP), Unemployment/Welfare/Other mandatory spending, and Interest on National Debt with a percentage of 21.41%,16.19%,14.63%,7.55%, 11.79%, and 9.10% respectively and the lowest spending is on Executive Office of the President and the Small Business Administration. This visualization is helpful for the U.S government as the spending information across different departments makes it easier for the U.S government to look into the budget request as per the expenditure.
Effectiveness of the visualization in conveying the story
According to me, the visualization is not able to convey the story very well, however, it can provide an idea that 31 U.S government departments are on the United States Federal Budget and it is differentiated by having different colors for the departments. By just looking at the chart I am not able to understand what are the percentages about or what is the comparison about. It just lists all the percentages and has a legend indicating what color is for which department. There are more than a dozen government departments represented in this graph of the U.S. government budget. Looking at the various color and size combinations for each pie slice becomes an unwelcome hurdle to understanding what this chart is trying to convey. I also observed in this pie chart that the percentages and the size of the pie in some of the representation of the U.S government departments were not complementing each other. There is also a repetition of the colors for representing the departments this makes it more confusing for the audience to know which department is having what percentage of spending. Since there are many departments, Some of the slices look like similarly-sized slices and some of the slices are too small. Those small slices are difficult to be read and identify distinct colors on them.
Whether the visualization is misleading in some way ?
I feel the visualization is misleading in some way due to the following reasons:
The visualization is might be misleading as there are many variables to present in this chart and the chosen way to present was a pie chart. This is an issue as it is difficult to straight away understand which proportions are significant and the legend is required to analyze this, also it is hard for people who are from the general public to analyze slices of a doughnut in this instance compared to if a bar graph was used and minor differences in percentages are difficult to make out in this visualization.
Another reason is the color scheme and the legend being used for the visualization. The legend correlates with the pie chart by using different shades of similar colors and some of the shades of the colors look alike and it’s make out the difference. The legend uses a small square of color to represent each department which is very difficult to see. Similar shades of color look almost identical. As this visualization is available to the public, it would be a good practice to account for those readers that may not be able to see such small variance in color and make visualizations with more contrast between colors.
The pie chart is often used to show percentage, where the sum of the sectors equals 100%. But in this Pie chart, I observed all the sectors did not equate to 100. So this leads to misleading the audience that the percentage calculated for each department is based on different underlying parameters.
Since the pie chart doesn’t have markings around their circumference, it looks harder to compare the departments of similar size, and the annotations for the smaller slice are written outside of the graph which looks very crowded for the audience to identify which percentage correlates with which slice and it’s misleading.
The title of the Chart " A pie chart representing spending by category for the US budget for 2007" makes it confusing for the audience to understand the visual as at the first glance of the graph we would not be sure what the categories are referring to.
How well the visualization holds up the visualization principles ?
The visualization is a pie chart and it is not able to hold the visualization principles we learned about very well. The graph does not provide the comparisons understandable, it looks cluttered. There is too much information provided in the visualization making it unable to convey the story successfullly and the focus is not maintained. There are many labels on the visualization and the labels are tiny, it is difficult to follow and read the information in the pie chart. Even though the data is labeled it is not well contrasted with color that will distinguish each part from the other. The colors used are repeated and they have a very slight difference to make sense that they are different from each other. This visual overstimulation detracts from the main point as it forces us to focus on every single category. For example, the Medicare department and Energy department have the same colors even when their proportions differ, and hence, the principle of Similarity does not hold good for this visualization. Also, because the chart has too much information the size of all the slices looks similar to each other. In this visualization, all the departments are placed next to each other for a better comparison and to look for similarities in such groups but the human eye does a poor job in judging angles and this makes it difficult to discern the values within pie graphs. Hence, the proximity principle does not hold good for this visualization. However, this visualization holds good for the principle of enclosure because each department appears to be set apart in a region that is distinct from the rest of the other departments. To conclude, I believe for information like this use of a different chart type would be more useful in showing the data.
How the visualization could be improved ?
The Visualization can be improved by using a bar chart. Instead of having color and size combinations for each pie slice, we may be better off sharing actual numbers without any visualizations at all. Since we have a considerable number of data points to summarize within one graph, we can overcome this challenge by using a bar chart when representing multiple data categories. While we’re at it, we could make sure to remove all unnecessary distractors such as too many colors and data labels. I also feel that for the U.S government or the public to look at the Budget request it would be more informative to include the actual spend data rather than having a percentage of expenditure. Lastly, we’ll have to sort the data in descending order to help users easily spot top spending categories. Using a bar or a column chart makes it easier to gauge exact values from bar lengths rather than slice areas or angles, especially since the bar chart naturally has one axis dedicated to value markings and no annotations would be necessary for a bar chart. It will help the visualization look less cluttered and more informative.
Improved version of the visualization
Why the improved version of the visualization is more informative ?
In the improved version of the visualization, I have used a bar chart and it is an information-dense visualization than the pie chart. My major concern was to show the proportions of spending across each U.s government department hence, I have used bars for each of the departments to give more clarity. The information flows clearly and tells a clear story of the expenditure across different U.S government departments. The bar graph is made horizontal so the variable names are easier to read, and the bars are ordered in descending order to easily spot top spending categories. The expenditure across each department can easily be compared and I have made sure the focus is maintained with the help of neat labeling and using the color-coding techniques. Further to reduce cluttering I have categorized the bars into two categories of spending type such as Mandatory and discretionary. “Mandatory spending,” also known as “direct spending,” refers to funds provided in laws other than appropriation acts, and “Discretionary spending” refers to the funds allocated by Congress to cover the administrative expenses of executive branch agencies, congressional offices, and agencies, and international operations of the government. Categorizing the bars in this way makes the visualization more informative aiding the U.S government in the decision of the Budget request. I have changed the percentage of spending to the dollar amount of spending across different departments and the title has been modified to provide clear information of what story is been told with the help of the visualization and axis labeling provides clear information about the variables. One glance at the graph provides all the information required. Hence, I think using a bar chart for such categorical data was more meaningful than using a pie chart.