Data Visualisation (Tableau Online): SEA Armed Conflicts
Data Source: The Armed Conflict Location & Event Data Project (ACLED), South East Asia, 2015 to 2020
1 Critiques and Suggestions
The original data visualization intended to reveal the spatio-temporal patterns of armed conflict in selected South-east Asia countries between 2015-2020.
Figure 1: Critic original DataVi
1.1 Clarity
| SN | Critiques | Suggestions |
|---|---|---|
| 1 | Axis for Armed Conflict Event by Type has different scale, it is intuitively misleading when comparing across. | Propose to use a single line chart so axis scale is aligned. |
| 2 | Event Type color legend is not used on the Armed Conflict Event by Type, which makes reader confused about the purpose. | Suggest aligning on the usage of color across different charts. |
1.2 Aesthetic
| SN | Critiques | Suggestions |
|---|---|---|
| 3 | When all countries are selected on the map, the boarder of countries is hard to tell. | Add one color layer for countries. |
| 4 | Count of Sheet1 does not have meaning and is redundant. | Remove it. |
1.3 Interactivity
| SN | Critiques | Suggestions |
|---|---|---|
| 5 | Selection of country does not reflect on the line chart, which does not give reader information about armed conflict event by type at country level. | Apply filter to multiple sheets. |
| 6 | Selection of event conflict does not reflect on the line chart. | Add “Action” to apply filter for same charts. |
2 Sketch of Proposed Design
Figure 2: Sketch of proposed design
On top of proposed solution to workaround critiques, we noticed some events are non-violence events with low fertility, one way to measure severity is by fertility, hence proposed to add key risk indicators, event and fertility counts for current year and Year over Year (YoY) percentage and breakdown by Event Group. Besides, propose to add a boxplot to illustration the distribution of fertility distribution by country.
Last, add actors in conflict to see the most common and deadly parties in armed conflict events.
3 Steps to Create Dataviz Using Tableau
3.1 Import data
The data source provided includes data from 2010 to 2020, while the original DataViz only includes 2015 to 2020, hence we removed 2010 to 2014 data from excel using column “Year”. Then, create one Tableau file and import data by clicking “Connect to Data” under Data pane and “Microsoft Excel” under To a File.
Figure 3: Import data
3.2 Reset Data Type and Edit Alias
Next, change “Year” from “Number (whole)” to “Date” by clicking the “#” symbol.
Figure 4: Change data type
Besides, change field “Interaction” to be “String” and split into two columns by right click the column and “Create Calculated Fields”, where for “int1” is the smaller actor type code in a conflict, and int2 is the larger actor type code in a conflict, when there is no 2nd type of actor, 0 is used. Formulas for “int1” and “int2” as below.
Figure 5: Create calculated field for actor type 1
Figure 6: Create calculated field for actor type 2
Figure 7: Create calculated field for actor type 3
Then right click “Int1” column to edit “Aliases…” with below changes.
Figure 8: Edit alias for actor type 1
Figure 9: Edit alias for actor type 2
3.3 Key risk indicators
To highlight key risk indicators of current year and YoY changes, we introduced to measure number of “Events” and “Fertalities”.
3.3.1 Create calculated fields
Calculated fields are created by clicking small triangle under Data pane.
Figure 10: Create calculated field
To monitor current year number of events, firstly, we need to create calculated fields to filter date.
Figure 11: Create calculated field, Recent Year
To compare with same duration in previous year, below field is created.
Figure 12: Create calculated field, Last Year Partial
Then, “Event|RY” is created to count number of events in recent year, and “Event | LYP” for same period last year.
Figure 13: Create calculated field, Event | RY
Figure 14: Create calculated field, Event | LYP
Then, YoY change percentage can be calculated.
Figure 15: Create calculated field, Event | % YOY
Lastly, below is created to represent positive or negative symbols.
Figure 16: Create calculated field, Event | % YOY | Down
Figure 17: Create calculated field, Event | % YOY | Up
Create “MIN(0.0)” to locate KPI at centre of field.
Figure 18: Create calculated field, MIN(0.0)
Also, group “Event Type” to “Event Group” by calculated field as below, based on ACLED_Codebook_2019FINAL.
Figure 19: Create calculated field, Event Group
3.3.2 Create KRI
Create a sheet by clicking the icon with vertical line on the bottom tab bar.
Figure 20: Create a new sheet
Double click the newly created tab to named as “Event total”, select and drag “Event Date” to Tooltip, “Event | RY”, “Event | % YOY”, “Event | % YOY | Up”, “Event | % YOY | Down” to Text.
Figure 21: Create sheet for number of events
Then click “Text” and then “…” button to format text to be displayed.
Figure 22: Edit number of events tab
Set “Events” as size, 11 and font, Tableau Book and bolded. “<SUM(Event | RY)>” as size-26, font-Tableau Bold. Also highlight YoY increase in red and decrease in green.
Figure 23: Edit label
Also remove zero-line and grid lines by selecting “Format” on the toolbar and “Lines”. Click arrows for “Grid Lines”, “Zero Lines” and “Axis Rulers” and set as “None”.
Figure 24: Edit line 1
Figure 25: Edit line 2
Click Tooltip under “Marks” pane to enter below text, so when hover around the chart, below will be presented.
Figure 26: Edit Tooltip
Duplicate this sheet by right click tab “Event Total” and choose “Duplicate”.
Figure 27: Duplicate sheet
Double click to rename as “Violent Event” and add “Event Group” as filter with “Violent events” selected.
Figure 28: Add Filter 1
Figure 29: Add Filter 2
3.4 Map
3.4.1 Create parameters and calculated fields
To enable map to be viewed by either event or fertility, add Parameter “View by” by clicking the triangle under “Data” pane and choose “Create Parameter…”, select Data type as “String” with “List” of values, “Event” and “Fertality”.
Figure 30: Create parameter 1
Figure 31: Create parameter 2
Then create calculated field “View by” to select field when “View by” parameter is selected.
Figure 32: Create calculated field, View by
3.4.2 Add map
Add “Country” and “Year” to Filters, “Country” and “Event Id Cnty” to “Detail”, “Event Group” to Colour, “AVG(Longitude)” as Column and “AVG(Latitude)” as Row.
Figure 33: Create map
Now we can see armed conflicts in South East Asia, but the boarder between countries is not clear, hence we add “AVG(Latitude)” again to Rows, and select “Country” as color, and synchronise them on same graph by choosing element type under “Marks” pane as “Map”, and right clicking second “AVG(Latitude)” and set as “Dual Axis”.
Figure 34: Add country layer in map
Figure 35: Apply dual axis in map
By doing so, the circles for individual events are hidden under the country layer, so bring it up by bringing second “AVG(Latitude)” to the font on the Rows.
Under “Marks” pane, click “Color” to “Edit Colors” and select Color Palette to choose desired one, e.g. “Blue-Teal” for map layer, click “Assign Palette” and “OK” to complete.
Figure 36: Edit color
Then select first “AVG(Latitude)” and update Tooltip as below.
Figure 37: Update Tooltip
3.5 Boxplot for fertility count in an event
3.5.1 Create parameters and calculated fields
Create Parameter “Jitter param” as below and calculated field “Jitter” to add jitter to box plot.
Figure 38: Create parameter, Jitter param
Figure 39: Create calculated field, Jitter
3.5.2 Create boxplot with jitters
Adding “Country”, “Jitter” to Columns, “Fertilities” to Rows, and “Event Group” to Color. Select “boxplot” under “Show Me”.
For color, reduce Opacity to “25%” to show the density of fertilities in one event.
Right click x-axis to remove jitters, and name chart title as “Fertalities Statistics by Country”
Figure 40: Create boxplot
Figure 41: Edit Opacity for jitters
Right click y-axis to set axis to start with “0”, “Logarithmic” scale and name as “Fertalities”.
Figure 42: Edit y-axis
Right click boxplot to “Edit” and set “Whiskers extend to” as “Maximum extent of the data”.
Figure 43: Edit boxplot
Add “Event Date” to Tooltip and update it as below.
Figure 44: Edit tooltip for boxplot
3.6 Association table for actors in conflicts
3.6.1 Create association table
Add “Int2” to Columns, “Int1” to Rows, “Country” to Tooltip, “View by” to Text, select “text tables” under Show Me.
Rename title as “Actors in Conflicts”, Tooltip as “
Figure 45: Create association table for actor types in conflicts
3.7 Line chart for breakdown by country, event type
3.7.1 Create calculated field
Create calculated field, “View by text”, to rephrase “View by” parameter selection.
Figure 46: Create calculated field, View by text
3.7.2 Create line chart for breakdown by country
Select “Year” as Columns, “View by” as Rows, “Country” as Color.
Update chart title as “Yearly Number of <Parameters.View by> by Country”, Tooltip as “<SUM(View by)> armed conflict <ATTR(View by text)> in
Right click x-axis and de-select “Show Header”.
Figure 47: Hide header for x-axis
Figure 48: Edit y-axis 1
Figure 49: Edit y-axis 2
Figure 50: Edit y-axis 3
Edit tooltip for line chart as below content.
Figure 51: Edit Tooltip for line chart
3.7.3 Create line chart for breakdown by event type
Similarly, create line chart with breakdown by event type like Section 3.7.2, by choosing “Event Type” to as colour differentiation.
3.8 Dashboard
3.8.1 Create dashboard
Create a dashboard by click the icon with table grid on the bottom tab bar.
Figure 52: Create dashboard
Enter Custom size with 1366px width X 1600px height by clicking triangles.
Figure 53: Edit dashboard size
3.8.2 Add Text objects and Sheets to dashboard
Select “Text” Object and move over to dashboard, to create title, footnote and etc..
Figure 54: Add text object
Figure 55: Create dashboard title
Then on the Sheets pane, move over relevant sheets to planned location, remove map sheet title as it is self-explained. Remove redundant filters.
3.8.3 Create Actions for interactivity
Select one sheet object in dashboard and click “Worksheet” on toolbar and choose “Actions…” to create actions. On the top up window, select “Add Action” and “Filter…”.
Figure 56: Select sheet and add action 1
Figure 57: Add action 2
Then another window pops up, for example, to add filter by country, first choose the source sheet with “Select” action and target sheet by clearing the selection will “Show all values”, click “Add Filter…” to add “Country” as filter for both source and target sheets.
Similarly add “event type” as filter action.
Figure 58: Add action 3
3.8.4 Final dashboard
Below shows final dashboard design.
Figure 59: Dashboard design
4. Insights
4.1 YoY events increased, and fertilities decreased
Number of events increase 10% YoY across three event groups, and fertalities decreased 34%, even though fertalities due to non-violent events increased 650%.
Figure 60: KRI
4.2 Violent conflicts and Demonstrations are most common event group
When View by Event is selected, we can see that violent events and demonstrations has high density.
Figure 61: SEA map by event
When View by Fertality is selected, we can see Philippines is where most fertilities happen due to violent armed conflicts.
Figure 62: SEA map by fertalities
4.3 Most severe conflict happened in Myanmar with 243 death
From boxplot below, we can see that one violent event happened on 28 August 2017 in Myanmar caused most death, 243.
Figure 63: Fertalities by
4.4 State Force, Political Militias and Rebel Group are involved in most conflicts
From Actors in Conflicts table, noticed that Political Militias, State Forces, Rebel Groups cause the most fertalities.
Figure 64: Actor type in conflicts table
4.5 Violence again civilians and protests are most common event type
From Yearly Number of Event by Event Type, we notice that Violence against civilians and protests are most common event type.
Figure 65: Violence against civilians, protests are most common conflict types
From fertality version, it is observed that Violence against civilians and battles cause most fertility.
Figure 66: Violence against civilians, battles are most deadly conflict types