Original


Source: NBC News (2018).


Objective

The primary intention of the first iteration of the data visualisation was to illustrate the levels of acceptance and dissatisfaction that people of different political affiliations had for President Donald Trump. People who are interested in American politics or people who consume news are likely to make up the majority of the target audience.

Audience

People who are interested in keeping up with political events in the United States make up the bulk of this visualisation’s target audience. This could include voters, political analysts, journalists, and anyone else who has an interest in the current political climate and how the public feels about President Trump.

Deconstruct/Critique

Cherry-picking, also known as “Design and Integrity, Avoiding Deception”: The data points that were presented in the first version of the bar graph had been carefully selected in order to highlight a certain narrative on Trump’s disapproval ratings. For instance, the data may have been chosen during a time period during which his approval ratings were lower the normal among various political affiliations. In addition, while they were producing the graph, they only questioned 100 people and eliminated those who had neutral opinions, which might have skewed the findings. In order to solve this problem, it is recommended to employ a bigger sample size as well as a wider variety of data points that are more representative in order to produce a more accurate portrayal of the public’s attitude. Adding and combining different voting sources to create and indpendent lable.

Omission of vital context (Design and Integrity, Storytelling with Data): Vital contextual information is missing from the original bar graph. There is no indication of demographic characteristics, and the graph gives the impression that the findings are based on the vote of Republicans for Trump, despite the fact that it only comprises one hundred votes. There is a lack of background information on the voters’ views, and there is no indication as to which party the voters were most strongly affiliated with. To overcome this problem, more information regarding the participants’ demographics, political affiliations, and opinions should be presented to the audience in order to provide them with a better understanding of the environment in which the data was collected.

Poor choice of chart type (Visual Perception and Colour, Grammar and Vocabulary): Due to the choice of chart type, the audience in the original bar graph finds it difficult to compare the degree of sentiment across different political affiliations (All registered voters, Republicans, Democrats, and Independents). This makes it difficult for the audience to compare the degree of approval (strongly approve, somewhat approve, somewhat disapprove, and strongly disapprove). The categories are not placed next to one another in a linear fashion, and there are no reference points available for making comparisons. As a consequence of this, any findings that are very close together might be visually deceiving, making it difficult to evaluate the data appropriately. In order to overcome this problem, a new form of chart that enables straightforward comparisons between various groups of people and degrees of emotion should be used.

Reference

NBC News. (2018). NBC/WSJ Poll: Trump stronger with base, weaker in the middle. Retrieved from https://www.nbcnews.com/politics/first-read/nbc-wsj-poll-trump-stronger-base-weaker-middle-n893601

Code

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

library(ggplot2)
library(dplyr)

# Create the data frame
sentiments_data <- data.frame(
  Category = factor(rep(c("Donald_trump", "Republican_party", "Democrats", "United_states_supreme_court", "Robert_Mueller", "ICE", "NATO", "Vladimir_Putin"), each = 5)),
  Sentiment = factor(rep(c("Very_Positive", "Somewhat_Positive", "Neutral", "Somewhat_Negative", "Very_Negative"), times = 8)),
  Percentage = c(28, 12, 9, 9, 43, 11, 24, 19, 20, 26, 10, 24, 23, 18, 25, 21, 29, 31, 12, 5, 18, 10, 20, 9, 13, 23, 15, 21, 16, 21, 20, 25, 28, 8, 5, 2, 3, 26, 19, 46)
)

# Create a new variable to represent the combined categories
sentiments_data <- sentiments_data %>%
  mutate(CombinedCategory = recode(Category,
    "Donald_trump" = "Trump, Republicans, Democrats",
    "Republican_party" = "Trump, Republicans, Democrats",
    "Democrats" = "Trump, Republicans, Democrats",
    "United_states_supreme_court" = "NATO, ICE, Supreme Court",
    "Robert_Mueller" = "Putin, Mueller",
    "ICE" = "NATO, ICE, Supreme Court",
    "NATO" = "NATO, ICE, Supreme Court",
    "Vladimir_Putin" = "Putin, Mueller"
  ))

# Create the line chart with facetting
p <- ggplot(sentiments_data, aes(x = Sentiment, y = Percentage, color = Category, group = Category)) +
  geom_line() +
  geom_point(size = 3) +
  facet_wrap(~CombinedCategory, scales = "free_y", ncol = 2) + # Change the number of columns to 2
  theme(axis.text.x = element_text(angle = 45, hjust = 1),
        plot.margin = margin(10, 10, 10, 10), # Adjust plot margins
        strip.text = element_text(size = 12), # Adjust facet label text size
        legend.position = "bottom", # Move legend to the bottom
        legend.title = element_text(size = 12), # Adjust legend title size
        legend.text = element_text(size = 10)) + # Adjust legend text size
  labs(title = "Sentiment Analysis of Public Figures and Groups",
       x = "Sentiment",
       y = "Percentage",
       color = "Category") +
  scale_color_manual(values = c("black", "darkblue", "darkgreen", "darkorchid", "darkorange", "gold", "darksalmon", "brown"))

# Display the chart

Data Reference

Wall Street Journal and NBC News. (2018). July 2018 WSJ/NBC News Poll. Retrieved May 1, 2023, from https://www.wsj.com/public/resources/documents/18570NBCWSJJuly2018FullPoll.pdf

Reconstruction

The following storyline addresses the most significant problems with the original.

The primary problems that were found in the original data visualisation were remedied by using a line chart that had facetting in the rebuilt version. The line chart makes it possible to compare various political affiliations and degrees of emotion in a way that is easier to understand, which makes it possible for the audience to more properly analyse the data. Through the use of faceting, related categories are grouped together, which not only provides additional context but also makes it simpler to recognise trends and patterns.

In addition to this, the new visualisation makes use of a bigger sample size, which provides a perspective that is more indicative of how the general population feels about President Trump. In addition to that, it provides additional information about the political connections and opinions of the participants, which enables the viewer to have a better understanding of the data.

As a result of resolving these shortcomings, the rebuilt data visualisation presents a picture of popular attitude towards President Trump that is clearer and more accurate, hence making it more informative and valuable for the audience it was designed for.