Before Adjustments

During a labor market lecture in my economics class, my professor presented a visualization which automatically caught my attention. The visualization was the one depicted above, sourced from the U.S. Bureau of Labor Statistics. It is aimed to illustrate the shifts in occupations within the United States, measured in the percentage of total enrollment. The main idea behind the image was to highlight the occurrence of “sectoral shifts” since the industrial revolution (a topic we were learning about, and which the image was supposed to help depict). The idea of sectoral shifts which looks to explore the bigger idea of frictional unemployment, where certain jobs no longer require the demand they once did, prompting individuals to acquire new skills to secure different highly demanded jobs.

Although with my understanding of the concept I was able to make out what the graph was trying to convey, one noticeable issue lies in its ability to effectively capture these changes. The similar colors for different occupations makes it challenging to make out what job is what. For instance, both the government and construction sectors are represented by the same green color, without any clear indication as to why this similarity exists. This lack of differentiation impacts the viewer’s ability to grasp the data. I originally thought that they chose the colors based on some sort of sub category, but this isn’t a very strong argument as “other professional jobs” and “forestry and fishers” share the same color but don’t necessary imply the same thing.

Creating a graph that’s visually appealing and informative is crucial, and shouldn’t require extensive searching for indicators. If done properly it should ease the viewer’s ability to interpret it. This can be achieved through clear titles, labels, color, accuracy of measurements and more. While some graphs may require prior knowledge for comprehension, this visualization should be pretty simple to understand, based on elements such as employment levels, years, colors and job titles/categories. However, the use of colors without explanation complicates understanding. Context is crucial in visualizations to ensure viewers grasp the data’s meaning accurately. Without good context, viewers may misinterpret the information presented. Similarly, the ineffective use of color can further effect comprehension by making it challenging to differentiate between categories or interpret trends. For example, the overly bright or saturated colors in this graph can also detract from the visual appeal and overall effectiveness of the visualization.

WHAT IMPROVEMENTS CAN BE APPLIED?

There’s several steps which can be taken in order to improve this design. For starters we would need to change the colors. Applying better colors as the current ones are a bit hard on the eyes and not appealing. Moreover, without clear explanations, the use of identical colors for different categories can lead to confusion. So, I would need to reassign them. By assigning colors based on labels or job categories enhances the clarity, which evidently makes it easier on the viewer to understand. Additionally, I would incorporate labels alongside each colored box, which can aid comprehension by providing immediate context, and reducing the need for the viewer to constantly refer back and forth to the labels on the right hand side, (The color changing step is particularly beneficial for managing the harsh depiction of colors and ensuring a smoother understanding of the data). The text is also a bit too small, so enlarging the text size of the labels on the side enhances readability, making it easier to comprehend the information presented. So, by carefully considering color choices, introducing immediate labels, and increasing text size, the visualization can be significantly improved in terms of clarity and effectiveness.

After Adjustments

As you may after a couple of adjustments this new graph is much easier to understand. I would love to add a title but, the original graph sourced from my class lecture didn’t provide one. So,if I were to create my own it would simply be “Percentage of Job Employment in the USA: 1910 V.S 2015.”

Note that the echo = FALSE parameter was added to the code chunk to prevent printing of the R code that generated the plot.