DataViz MakeOver 3

A critique and proposed makeover of the data visualisation of South-east Asia Armed Conflict Analysis published on Tableau Public. The data (South-East Asia) used for the visualisation could be obtained from The Armed Conflict Location & Event Data Project (ACLED).

Published

March 14, 2021

Citation

Wong, 2021

Overview

The original visualisation is as shown below:

A. Critiques

Clarity

  1. The y-axis titles of the line plots under Armed Conflict Event by Type of different event types are named as “Count of Sheet1”. It is unclear to the reader as to what is “Sheet1”.
  2. Data source is not added. The credibility of the visualisation is questionable.
  3. The data range used differ across countries. For example, the line graphs of Malaysia show only 2018 to 2020 information. This can be misleading as the title of the map continues to show “2015-2020”.
  4. There are different shapes observed for the same event type/ color. It is unclear what the different shapes represent since there is no legend provided.

Aesthetics

  1. With reference to the Armed Conflict Location & Event visualisation on the right, several countries have high concentration of events points with certain events overlapping the others. Readers are not able to easily identify the other events in the same area. Examples of places or spots with high concentration of events include the whole of the Philippines, northern part of Rakhine State of Myanmar, and Pattani, Yala, and Narathiwat provinces at the southernmost of Thailand.
  2. The header of the line plots should not be “Event Date” as the default x-axis shows only the years.
  3. The word “Event” on the title of the map does not change when a specific Event Type is selected.
  4. The data points on the map have white borders that are redundant.

Interactivity

  1. Country filter is provided to allow users to focus on country of their interest.
  2. Event Type filter highlights the corresponding line graph and points on the map by lightening the colours of the non-selected ones. This enables users to quickly identify the graph and event they would like to focus on.
  3. The hover-over information/ tooltip of the individual points on the map does not provide much useful information.

B. Proposed Improvement and Data Visualisation

Clarity

  1. The y-asix title of the line graph should be amended from “Count of Sheet1” to “Number of Events”.
  2. Include the data source used for the visualisation.
  3. The tooltip of the filled map should have irrelevant information removed and with useful information added.
  4. The tooltip of the symbol map should include other information such as sub event type and notes for each event point.

Aesthetics

  1. Align the x and y-axis of all the line plots of different event types.
  2. Merge all the events into one single filled graph so that users could compare the severity or frequency of different events within the same country.
  3. The filled map is to be coloured by the same colour of different gradient to show the difference in the number of armed conflicts.
  4. Place a symbol map beside the filled map to show a similar visualisation with points coloured by event type and sized by number of fatalities.
  5. For the symbol map, the points are to have slight transparency so have to reveal the “hidden” event points overlapped by others

Interactivity

  1. When a country is selected on a filled map, the view should be zoomed in to focused on the country only. The corresponding symbol map should also be focusing on the country selected.
  2. The tooltip should showed more information provided in the dataset, as mentioned under Clarity. It shoudl have table of data if necessary.
  3. The titles of the visualisations should be updated interactively as users make changes to the filters.

Proposed Data Visualisation

C. Data Visualisation in Tableau

The interactive dashboard created could be found here

D. Preparation of Data Visualisation

1. Loading and Preparation of Data

2. Creating Filled Map

3. Creating and Editting Tooltip for Filled Map

4. Creating Symbol Map

5. Creating Line Graphs for Number of Armed Conflicts

6. Creating Line Graphs for Fatalities

Creating the Dashboard

E. Observations

  1. In the past 6 years, from 2015 to 2020, with the exception of Indonesia, Philipiines, and Thailand, the number of armed conflicts in the region stayed pretty much constant in general with no improvement or escalation. Both Indonesia and Thailand saw a stark increase in the number of Protests, while the Philippines saw a dip in the number of Violence against civilians.
  2. With the exception of Myanmar and the Philippines, most of the countries in the region had low fatalities despite multiple armed conflicts of different types.

  1. From the colour gradient and the hover-over information, it could be observed that the Philippines has the most number of conflicts in the last 6 years, followed by Myanmar, Indonesia, Thailand, Cambodia, Malaysia, Vietnam, and Laos.

  1. Zooming in on the Philippines by selecting the filled map on the left, and then observing the points on the symbol map on the right, it could be observed from the size of the points that while the country has more Violence against civilians than Battles, it has relatively lower fatalities per event than Battles.

  1. The area of the neighboring countries surrounding Singapore are rather peaceful in the past 6 years. A small number of Protests broke up but were of little to no fatality.

Distill is a publication format for scientific and technical writing, native to the web.

Learn more about using Distill at https://rstudio.github.io/distill.

Footnotes

    Citation

    For attribution, please cite this work as

    Wong (2021, March 15). Dylan's DataViz Blog: DataViz MakeOver 3. Retrieved from https://dylanwong.netlify.app/posts/2021-03-15-dataviz-makeover-3/

    BibTeX citation

    @misc{wong2021dataviz,
      author = {Wong, Dylan},
      title = {Dylan's DataViz Blog: DataViz MakeOver 3},
      url = {https://dylanwong.netlify.app/posts/2021-03-15-dataviz-makeover-3/},
      year = {2021}
    }