Click the Original, Code and Reconstruction tabs to read about the issues and how they were fixed.

Original


Source: howmuch.net (2019).


Objective

  • The objective of the data visualisation was to details how the average Americans allocate consuming budgets in 2017, and therefore analyse the economic trends. The data visualisation was built based on the U.S. Bureau of Labor Statistics data, which included the average annual expenditure in 2017, percentage distribution of total spending by major categories. The selected categories included Housing, Transportation, Food, Personal insurance and pension, Healthcare, Entertainment, Cash contributions, Apparel and services, Education, and Other expenses.

  • The target audience of this data visualisation was general public who interested in American household daily expenditure. They can have an idea of the how much was spend each year per person in each particular categories, and understand the percentage of daily expense.

The visualisation chosen had the following three main issues:

  • Inappropriate choice of visualisation As the data visualization is designed to compare the percentage of personal expenditure by categories in 2017, the original visualization may not convey the information properly because it is difficult to compare the size of each part and difficult to sort. For example, looking at the graph, it cannot visually tell the percentage difference between food (12.9%) and transportation (15.9%). It could have a better way to visualize this dataset by using bar chart, which is more clear to compare the differences between different categories.

  • Inappropriate use of colors Data visualization aims to differentiate expenditure categories by color. However, this visualization uses too many approximate colors, and we can’t tell how the red categories are related to the purple categories. It is also not sorted by shade of color, which makes it difficult to understand the information. The colors used are not suitable to conveying the intended message.

  • Unnecessary elements The visualization looks very eye-catching, but it presents some unnecessary features, such as using icons and complex colors representing each category . These should be avoided as they may confuse the audience. Text is easier to read than colors and icons. Text should be shown in the figure, rather than using icon.

Reference

Code

The following code was used to fix the issues identified in the original.

#install library
library(ggplot2)
library(readr)
library(tidyverse)

#enter data
spending <- read_csv("spending.csv")
view(spending)

#select necessary data which relevant to the original visualisation
spending2 <- spending [c(1,9)]
spending2 <- na.omit(spending2)

#sort 2017 data
spending2 <- spending2 [order(spending2$`2017`,decreasing = T),]
view(spending2)

#add relevant information
spending2 $ `Percent(%)`<-c(33.1,15.9,12.9,11.3,8.2,5.3,3.8,3.1,3.1,2.5,0.9)
view(spending2)

#plot the graph
p1 <- ggplot(data=spending2,aes(fill="Percent(%)",y="item",x="2017"))
p1 <- p1 + geom_bar(position = "fill",start="identity",colour = "turquoise3")+
  scale_y_continuous()+
  labs(title = "Consumer Spending in the United States 2017", 
       y="Category",x="Average Annual Personal Expenditures by Category")+
  theme_grey()

Data Reference

Reconstruction

The following plot fixes the main issues in the original.