Getting Ready for Halloween 2023

In this web report, we will thoroughly review the data visualization titled “Halloween Tracker” by the Reddit user “LozTriforce.” The visualization presents data on the number of kids coming to their home in Washington State for Trick or Treating during Halloween since 2009. Our primary focus will be to identify weaknesses in the original data visualization and propose effective solutions to address them.

To facilitate a comprehensive analysis, we have divided the report into three sections:

Original: In this section, we will critically examine the weaknesses observed in the original “Halloween Tracker” data visualization. We will discuss issues related to data integrity, visual design, color choices, and any potential ethical concerns.

Code: Here, you will find the code used to create the new data visualization. The code provided in this section will address the problems identified in the original visualization, offering a solution to improve its effectiveness.

Reconstruction: This section will showcase the new and updated data visualization created using the code from the previous section. The updated visualization aims to resolve the weaknesses found in the original version, enhancing its clarity and overall impact.

By exploring each of these sections, you will gain a comprehensive understanding of the weaknesses present in the “Halloween Tracker” data visualization and how we effectively addressed them through the creation of the updated visualization. Our objective is to deliver valuable insights into the visualization’s potential and empower future data visualizations to effectively communicate information with precision and visual appeal.

Click the Original, Code and Reconstruction tabs to explore each section.

Original Data Visualisation

Objective of the original data visualisation: The original data visualization serves the purpose of preparing residents in Washington State for Halloween trick-or-treaters in 2023, with a specific focus on optimizing candy purchases. The visualization aims to inform the audience about the notable decrease in the number of kids coming for Halloween due to the effects of COVID-19. By understanding this trend, the target audience, which likely includes homeowners, candy suppliers, and local communities, can make informed decisions on how many candies to purchase for the upcoming Halloween. The visualization intends to empower the audience to avoid excessive candy leftovers from the previous year by accurately anticipating the expected number of trick-or-treaters for the current year.


Source: [OC] We keep track of how many kiddos come for full-sized candy bars on Halloween (2023).


Reference:

loztriforce. (2023, July 25). [OC] We keep track of how many kiddos come for full-sized candy bars on Halloween [Data visualization]. Reddit. Retrieved Month Day, Year, from https://www.reddit.com/r/dataisbeautiful/comments/158rk0o/oc_we_keep_track_of_how_many_kiddos_come_for/

Analysis of the Original Data Visulisation:

1. Seeing Impairment Color - Red on Green: Using red and green color combination in data visualization can be problematic for individuals with color vision impairments, particularly those with red-green color blindness. The choice of colors that are challenging to distinguish for such individuals may lead to confusion and misinterpretation of the data. Red-green color blindness is one of the most common types of color vision deficiency, affecting how people perceive these colors. Consequently, viewers with this impairment may struggle to differentiate between data points represented in red and green, hindering their ability to accurately understand the information presented in the visualisation.

Reference:

Datylon. (2022). Data Visualization for Colorblind Readers. Retrieved from https://www.datylon.com/blog/data-visualization-for-colorblind-readers.

Colour Blind Awareness. (2023). Home. Retrieved July 26, 2023, from https://www.colourblindawareness.org/

2. Numbers Do Not Align to Scale: Misalignment of numbers to the scale in the data visualization can significantly impact its accuracy. When data points are not precisely positioned according to the scale, it may lead to misinterpretation of trends and relationships between data. For instance, if data points are not accurately plotted along the scale, the viewer may perceive an erroneous increase or decrease in values, potentially leading to incorrect conclusions about the data trends. Proper alignment is crucial to ensure that the visual representation accurately reflects the underlying data, enhancing the reliability of the visualization.

References: “Data Visualization Checklist” by Ann K. Emery: http://annkemery.com/dataviz-checklist/ “Datawrapper Academy: Scaling” by Datawrapper: https://academy.datawrapper.de/article/117-scaling

3. Number of the Y-axis Scale Should Start at 0: In a bar chart, not starting the y-axis scale at zero can distort the perception of data proportions. This decision may exaggerate differences between data points, making them appear more significant or significant than they actually are. By not adhering to the “start at zero” principle, the visualization may mislead the viewers about the relative magnitudes of the data, impacting their understanding of the data’s true values and trends.

References: Lotto, B. (Speaker). (2017, June). Optical illusions show how we see | Beau Lotto [Video]. TED. https://www.youtube.com/watch?v=mf5otGNbkuc

Code

The following code was used to fix the 3 problems identified in the original visualisation.

# Load the libraries required to fix the problems
library(ggplot2)
library(readxl)

# Loading the data from the orginal visualisation
halloween_data <- read_excel("HalloweenTracker.xlsx")

# Bar chart with improved color choices and subtitle
new_plot <- ggplot(halloween_data, aes(x = as.factor(Year), y = Kids)) +
  geom_bar(stat = "identity", fill = "skyblue") +
  labs(title = "Halloween Tracker - Number of Kids Trick or Treating",
       subtitle = "Tracking the number of kid Trick or Treating since 2009 in Washington State. Note: 0 years are due to COVID",  
       x = "Year",
       y = "Number of Kids Trick or Treating") +
  theme_minimal() +
  theme(plot.subtitle = element_text(size = 8, face = "italic", hjust = 0.5))

Reconstruction

Here is the updated data visualization representing the “Halloween Tracker” data. The bar chart displays the number of kids trick or treating in Washington State for each year since 2009, addressing the issues identified in the original visualisation.

## Source: We keep track of how many kiddos come for full-sized candy bars on Halloween (2023).

Reference:

loztriforce. (2023, July 25). [OC] We keep track of how many kiddos come for full-sized candy bars on Halloween [Data visualization]. Reddit. Retrieved July 26, 2023, from https://www.reddit.com/r/dataisbeautiful/comments/158rk0o/oc_we_keep_track_of_how_many_kiddos_come_for/

Reconstruction of the original data visualisation: In the new visualisation, the following changes have been made to address the problems of the original visualisation:

1. Color Choice:

The new bar chart uses a different color scheme to address the issue of using red and green colors together, which can be problematic for individuals with color vision impairments. Instead of the original color scheme, the new bar chart uses a “skyblue” color for the bars, providing a visually distinct color that is easily distinguishable from red and green.

2. Alignment of Numbers to Scale:

The new bar chart ensures that the data points are accurately positioned along the scale on the y-axis. This alignment is achieved by using the as.factor(Year) function for the x-axis variable, which converts the “Year” variable into a factor. This conversion ensures that the data points are plotted with equal spacing along the x-axis, even if the years are not consecutive, improving the alignment of the data with the scale.

3. Starting Y-Axis Scale at 0:

The new bar chart follows the best practice of starting the y-axis scale at 0 for bar charts. By default, ggplot2 starts the y-axis scale at 0 for bar charts, which helps prevent distortions in data proportions and exaggeration of differences between data points. The new visualization adheres to this best practice, enhancing the accuracy of the representation of data.

Overall, the changes in the new bar chart aim to improve color choices, accurately represent data points along the scale, and follow best practices for bar chart design. These improvements enhance the clarity, accuracy, and interpretability of the data visualisation, providing a better and more effective representation of the “Halloween Tracker” data. This makes it easier for the audience prepare to for the Washington State Halloween Season 2023 being ready with just enough candy and forecasting how many trick or treators may arrive.