DataViz Makeover 3
This dashboard surfaces the most common and deadliest armed conflicts in Southeast Asia, from 2010 to 2020.Visit the dashboard here. 
This dashboard surfaces the most common and deadliest armed conflicts in Southeast Asia, from 2010 to 2020.Visit the dashboard here.
On 1 February 2021, the world woke up to news of a coup in Myanmar. Innocent lives were ended and economic activity virtually brought to a standstill. The reality of political violence, societal harm and economic damage that armed conflict leaves in its wake was once again brought to the forefront of our collective minds.
In light of these developments, this dashboard visualises armed conflict in Southeast Asia from 2010 to 2020, with the objective of surfacing the most common and deadliest types of conflict.
This visualisation makes use of data provided by The Armed Conflict Location & Event Data Project (ACLED).
This is a visualisation created to reveal the saptio-temporal patterns of armed conflict in selected Southeast Asian countries between 2015 to 2020. 
The remade data visualisation was created using Tableau Desktop 2020.4.2 and can be viewed on Tableau Public.
| S/n | Critique | Suggestions for improvement |
|---|---|---|
| 2 | The visualisation is clear in presenting the event types (types of conflict). It also allows the user to select which country to focus on. However, this is very limited information and does not show any other information that might be useful in such analysis. | Allow the user to drill down into deeper information such as event sub types and number of fatalities specific to each region within a country. |
| 1 | The 6 line graphs do not have uniform ranges in the y-axis. This makes it confusing for the user and difficult to compare between different event types. | Redesign the line graphs such that the y-axis ranges are uniform. |
| S/n | Critique | Suggestions for improvement |
|---|---|---|
| 1 | The dots in the map do not provide any further information other than the location of the event. | Allow the size of the dots to represent the number of fatalities. |
| 2 | The labels in the line graph are superfluous. If removed, the size of the graph can be increased for better ease of viewing. | Remove the labels in the line graph and use colours to represent instead. |
| S/n | Critique | Suggestions for improvement |
|---|---|---|
| 1 | The visualisation does not allow selection of specific months or years for deeper analysis. This limits the usability of the visualisation. | Include an interactive feature that allows the user to select a specific month and year to focus in to. |
| 2 | The visualisation does not allow the user to select more than one country at a time. Such features might be useful to the user for comparison and analysis. | Include a feature to allow the user to select more than one country at a time. |
The proposed visualisation above provides more in-depth analysis such as the proportion of each event type and sub type, as well as a comparison of the number of fatalities of each event type.
The visualisation allows the user to select the specific month and year for deeper analysis.
the visualisation allows the user to select more than one country for presentation, for more useful analysis. The tree map of sub event types also allows for this same feature of selecting multiple sub event types for analysis, instead of only a single type each time.
The visualisation makes use of colour to replace labels, allowing for more intuitive associations, as well as to free up more space to allow the charts to be larger.
Step 1: Connect to the data set. 
Ensure data types are correct. Step 2: Change Region to Country/Region type. 
Step 3: Change Admin1 to State/Province type. 
Step 4: Change Admin2 to County type. 
Step 5: Change Admin3 to City type. 
Step 6: Create a calculated field Inter1(recoded). 
Step 7: Enter the following formula to show the specific terminology instead of the code. 
Step 8: Create a calculated field Inter2(recoded). 
Step 9: Enter the formula below. 
Step 10: Create a calculated field Interaction(recoded). 
Step 11: Enter the formula shown to show the exact interaction instead of only the interaction code. 
Step 12: Open a new Sheet. Drag Country, Admin1, Admin2, Admin3 onto Region to create a hierarchical group of geographical data. 
Step 13: Drag Country to the middle of the work pane and select symbol maps from the Show Me menu. 
Step 14: In the Marks pane, click on the “+” sign in front of Country to expand it until Admin3 is shown. 
Step 15: Remove Admin1 and Admin2. 
Step 16: Drag Event Type on to the Color Mark in the Mark pane. 
Step 17: Edit the transparency and border settings for better visual clarity. 
Step 18: Drag Fatalities on to the Size Mark in the Mark pane. Ensure it is measured by SUM to show the total number of fatalities. 
Step 19: Adjust the size of the circles for better clarity. 
Step 20: Drag the Country field to the Filters pane. When the settings dialogue pops up, select Use all. 
Step 21: Right-click on Country and select Show Filter. 
Step 22: Click on the drop-down arrow on the Country filter and select Single Value (dropdown). 
Step 23: Do the same for the Event Type field. 
Step 24: Drag Event Date to the Pages pane twice. The first time will automatically set to YEAR; the second time will automatically set to QUARTER. 
Step 25: Right-click on the QUARTER field and select Month. 
Step 26: Drag Admin1 to the pane and expand it to show Admin2 and Admin3. Drag Fatalities to the Size Mark, and Event Type to the Color Mark. 
Step 27: Drag Country field on to the map. The dialogue Add a Marks Layer will appear. Drop the Country field on the icon to create a new Mark Layer. 
Step 28: Drag Fatalities field on to the Color layer in the Marks Layer pane. 
Step 29: Right-click on SUM(Fatalities) to edit the colour settings. 
Step 30: Adjust the colour settings as shown. 
Step 31: Click on the Tooltip Mark in the Event Type Mark pane and edit the Tooltip as shown. 
Step 32: Click on the Tooltip Mark in the Country Mark pane and edit the Tooltip as shown. 
Step 33: Rename Sheet as Map of armed conflict in Southeast Asia. 
Step 34: Right-click on Sheet title and select Hide Title, as this is not necessary for the dashboard. 
Step 35: The final design of this sheet is as shown. 
Step 36: Open a new Sheet. Drag the Region and Event Date fields to Columns, and Event Type and Fatalities fields to Rows. 
Step 37: Drag the Event Type and Country fields to the Colors Mark. 
Step 38: Drag Event Date to the Filters Pane, and select Months. Click Next. 
Step 39: Select Use all. Click OK. 
Step 40: Select the following settings for the Country field. 
Step 41: Right-click on the y-axis of any of the 6 charts and un-select Show Header. 
Step 42: Right-click again on the y-axis and select Edit Axis. 
Step 43: Set the following settings. 
Step 44: Right-click on the x-axis label and un-select Show Header. 
Step 45: Click on the Tooltip Mark in Mark pane and edit the Tooltip as shown below. 
Step 46: Rename Sheet as Trending of event types. 
Step 47: Edit the Sheet title as shown. 
Step 48: The final design of this sheet is as shown. 
Step 49: Open a new Sheet. Drag Sub Event Type to the Marks Pane three times: once over to the Size Mark, a second time to the Color Mark, and a third time to the Text Mark. 
Step 50: Right-click on the Color Sub Event Type and change the measure to Count. 
Step 51: Do the same for the Size Sub Event Type. 
Step 52: Right-click on the Color Sub Event Type again and change Quick Table Calculation to Percent of Total. 
Step 53: Do the same for Size Sub Event Type. 
Step 54: Drag Event Type, Event Date (by Year), Event Date (by Month), and Country to the Filters pane. Right-click on each of them and select Show Filter. Customise the filters as shown on the right-hand side pane: Country and Event Type as Multiple Values (dropdown), YEAR(Event Date) and MONTH(Event Date) as Single Value (slider). 
Step 55: Click on the small drop-down arrow on % of Total CNT(Sub Event Type) and select Edit Colors…. 
Step 56: Customise the colour settings as shown. 
Step 57: Click on the Tooltip Mark in Mark pane and edit the Tooltip as shown. 
Step 58: Rename the Sheet as: Proportion of sub event type. 
Step 59: Edit the Sheet title as shown. 
Step 60: The final design of this sheet is as shown. 
Step 61: Open a new Sheet. Drag Event Type to the Color Mark, and a second time on to the Size Mark. 
Step 62: Right-click on the Size Event Type and change the measure to Count. 
Step 63: Right-click again on the Size Event Type and change Quick Table Calculation to Percent of Total. 
Step 64: Drag Event Date (by Year), Event Date (by Month), and Country to the Filters pane. Right-click on YEAR(Event Date) and MONTH(Event Date), and select Show Filter. Customise the filters as shown on the right-hand side pane: YEAR(Event Date) and MONTH(Event Date) as Single Value (slider). 
Step 65: Click on the Tooltip Mark in Mark pane and edit the Tooltip as shown below. 
Step 66: Rename the Sheet as Pie chart of event types. 
Step 67: Edit the title as shown. 
Step 68: The final design of this sheet is as shown. 
Step 69: Open a new Sheet. Drag Event Type to the Columns Pane, and a second time on to the Rows Pane. 
Step 70: Drag Event Type to the Color Mark. 
Step 71: Drag Event Date (by Year), Event Date (by Month), and Country to the Filters pane. 
Step 72: Right click on the column field label and select Hide Field Labels for Columns. 
Step 73: Right-click on the y-axis and un-select Show Header. 
Step 74: Right-click on any of the bars and un-select Show Header. 
Step 75: Click on the Tooltip Mark in Mark pane and edit the Tooltip as shown. 
Step 76: Edit the title as shown. 
Step 77: The final design of this sheet is as shown. 
Step 78: Open a new Dashboard. Drag the following Sheets into the Dashboard: Map of armed conflict in Southeast Asia, Trending of event types, Proportion of sub event type, Pie chart of event types, Fatalities by event type and arrange them and the filter controls as shown. 
Step 79: Set the size of the dashboard to Automatic. 
Step 80: At the Country filter control, click on the small drop-down arrow and select Apply to Worksheets, Selected Worksheets…. 
Step 81: Select the Sheets as shown. 
Step 82: Perform the above 2 steps for filter controls Year and Month but do not select Trending of event types. 
Step 83: For the Conflict Type filter control that filters only the tree map, select Only This Worksheet. 
Step 84: Edit the title of the dashboard as shown.
Step 85: The final dashboard design is as shown.
This are even more fatalities than battles between armed and organised actors. This immediately raises a concern that the ordinary civilians seem to be at higher risk of losing their lives amongst all types of violence. 
From the year-on-year trend of conflict types, it can be seen that there was a large spike in the number of fatalities in 2016. Upon closer examination, this seems to have been contributed in large part from The Philippines. HOwever, we should also take into consideration that data from the Philippines prior to 2016 is not available in this data set. 
If any solace can be found, it is in the fact that all conflict types have generally been on the decline.
This can be seen from the year-on-year chart which shows that there were peaks in the number of fatalities resulting from battles in 2011, 2015, and 2019. This seems to coincide with the dissolution of the military junta in 2011, the elections in 2015, and fighting between the Myanmar army and the ethnic Rakhine-backed Arakan army in 2019. It is noteworthy that this seems to be a 4-year cycle, which may hint at a systemic issue to be mitigated. 
The Philippines had the highest number of fatalities among all Southeast Asian countries from 2010 to 2020, as can be seen from the deeper red hue on the map. 
Focusing in to The Philippines, the bar chart shows that from 2016 to 2020, the highest number of fatalities are due to violence against civilians. 
The pie chart corroborates this as it shows that the most common conflict type was violence against civilians. 
The tree map shows that attacks were the most common conflict sub-type, accounting for almost 60% of all conflict sub-types. Attacks are defined in the ACLED Codebook as “when civilians are targeted with any violence by an organised armed actor”. 
Turning our focus to the next conflict type that accounts for the most fatalities - battles - the year with the most number of battles was 2017. Zooming in to 2017, we see that The Philippines accounts for the most number of battles in that year. 
Focusing in on The Philippines, the battle with the most fatalities for that year was in Bangsamoro Autonomous Region in Muslim Mindanao, Lanao del Sur, Marawi - with 1,192 fatalities. 
Even so, violence against civilians still ‘topped the charts’ in terms of number of events, and number of fatalities, corroborating with observation 1.
Protests make up almost 30% of all conflict types from 2010-2020, but accounts for only 108 fatalities. The most fatalities coming from a conflict in Dien Bein, Moung Nhe, Viet Nam. 
From this, we can see that protests are the least deadly in terms of fatalities. Contrasting this with violence against civilians which has the highest number of fatalities, we can reach the conclusion that civilians involved in armed conflict can be vulnerable to violence but can also generally partake safely in peaceful protests. This may advise how civilians and rights groups can strategise their activities to make their voice heard in a way that minimises - if not totally elliminates - harm to civilians.
Thanks for visiting my blog!
This post is a data visualisation assignment for the MITB programme at the Singapore Management University. 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.