Assignment 2

Deconstruct, Reconstruct Web Report

Original

Source: U.S. Census Bureau (2019).

Objective

The original objective of the visualisation is to show the percentage of each U.S. states population without any health insurance. This is achieved by using a heat map showing each state with a colour based on the percentage of the population uninsured, this is confusing as the designer has chosen visuals over function.

The target audience for this visualisation could be insurance companies, looking to decide where they could spend marketing budget to reach potential new customers. It could also act as an incentive for the general public to consider their current insurance status.

The visualisation chosen had the following three main issues:

• Issue 1: Choice of visualisation style is attempting more of an aesthetic approach rather than a functional approach. That is, it designed to be resemble a map of the United States The visualisation draws the viewers eye, but the choice to make it eye catching overpowers the meaning.

If we look at the trifecta check for this visualisation the question is clear, however the chart can be difficult to decipher and what the data is saying is muddled across the chart, with no straightforward patterns available. The result would be DV of the 8 possible critiques for data visualisations.

• Issue 2: The size of each data point is equal, even though the data is not. This is deceptive as on first glance it appears that each state is equal. You need to read into the chart to understand the differences in colour and what the percentage they represent.

• Issue 3: The colours used are not ideal for colour blindness. I ran the visualisation through Coblis and the results showed that for various types of colour blindness the visualisation is quite hard to interpret what is happening in the visualisation. Another issue with the colours is the gradient used makes the variation between the highest and lowest rates of the population without health insurance seem more extreme than it actually is, there is only a variation of 15.4%.

Reference

• U.S Census Bureau. (2019). Health Insurance Coverage in the U.S. - Percentage of Population Without Health Insurance. Retrieved March 28, 2022, from howmuch.net website: https://howmuch.net/articles/health-insurance-coverage-in-the-us

Code

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

library(dplyr)
library(ggplot2)

Insurance <- read.csv("Insurance.csv")

Insurance$State <- Insurance$State %>%
  factor(levels = Insurance$State[order(-Insurance$Percen)])

p <- ggplot (data = Insurance, aes(x = Percen, y = State))
p <- p +geom_col (colour = "#FFFFFF", fill = "#7393B3") + 
  geom_text(aes(label = Percen),hjust = 1, nudge_x = -.5, size = 3, colour = "black") +
  labs(title = "Health Insurance Coverage in the US:", 
       subtitle = "Percentage of Population Without Health Insurance",
       x = "% of Uninsured Citizens",
       y = "State")

Data Reference

• U.S Census Bureau. (2019). Health Insurance Coverage in the U.S. - Percentage of Population Without Health Insurance. Retrieved March 28, 2022, from howmuch.net website: https://howmuch.net/articles/health-insurance-coverage-in-the-us

Reconstruction

The following plot fixes the main issues in the original.