Political Violence and Protests in South East Asia

This was written as part of the requirements for the Visual Analytics module for MITB.

For this DataViz Makeover, I have used data from The Armed Conflict Location & Event Data Project (ACLED), a centralized hub for data on all reported instances of political violence and protests around the world.

1. The original visualisation

Before making over the selected visualisation, it is important to have a clear understanding of the context of the visualisation and its key takeaways which are dependent on:

• Who (target audience): Practitioners, researchers, journalists, and governments.

• What (intended meaning): To reveal the spatiotemporal pattern of conflict in selected South-east Asia countries.

• Why (desired effect): To help users (practitioners, policy makers, etc) decide where, how and when to allocate resources

2. Critiques and suggestions for current visualisation

My critique will be based on Ben Jones data visualisation’s four quadrants of clarity and aesthetic, as well as interactivity of the visualisation.

Ben Jones defined clarity as “Does it make sense” and aesthetics as “Does it look good”.

Based on this concept with its 4 quadrants, I have classified this original visualisation to be in the III quadrant as the current visual only tells us the event type location and general trend across the 5 years, with the 2020 data being incomplete and thus misleading.

2.1 Clarity

2.2 Aesthetics

2.3 Interactivity

3. Proposed Design

3.1 Sketch

Notes

1. Overlay instructions on interactivity.
2. Overlay Country by event type chart per country chart.
3. Use red shades for violent events, blue shades for non-violent events.
4. Size of map bubbles to reflect number of fatalities.
5. Point symbol map - each bubble represent an event.
6. Map tooltip to include event location, date, notes and reported fatalities.

3.2 Advantages of Proposed Design

1) Visual provides good overview of the location, severity and frequency of events
At first glance, users are able to have a sense of the frequency (via number of bubbles) and severity (in terms of fatalities) of the events (via size of bubbles) of respective countries.

2) Tooltip provides event details for deeper analysis
Users are able to investigate an event further by hovering over a point on the map. Details in tooltip includes event location, date, notes and reported fatalities.

3) Interactive time series chart - Event/fatalities over time
Users are able to customise viewing according to needs - (i) events type count / fatalities by event, (ii) by date level - week / month / year, (iii) time period to observe and (iv) overlay total fatalities on top of the chart. From here users are able to identify a specific year/month/week for deeper analysis if needed.

4) Meaningful colour palatte
Red shades represent violent events, while blue shades are non-violent events. Colours are consistent throughout the dashboards.

5) Map - year filter
Shows events by selected year. Users are able to select a specific year for analysis.

6) Instructions on how to use the interactive visual
With this instructions, users are able to solely focus on analysing conflict.

7) Main groups involved
Users are able to identify the main groups involved in the conflict.

8) Country and event type filters
User are able to view selected country(ies) and event type(s) on the map for comparison.

4. Final visualisation

4.1 Dashboard

Due to limitation of webpage width, visualisation may appear compressed. Hence to enjoy the full experience of this visualisation, please view in Full Screen mode (bottom right of the visualisation). This visualisation is also available on Tableau Public.

4.2 Main observations

1) Overview of political violence and protests in SEA 2019

In year 2019, majority of Philippines’ and Myanmar’s events were violent conflicts (Battles, Violence against civilians and explosions).
Whilst Indonesia experienced a significant number of protests.
Thailand experienced a fairly even mix of all events.
The remaining countries – Malaysia, Cambodia, Vietnam and Laos experienced a small number (relative to the region) of protests.

These violent conflicts tapered since 2018 but with more protest on the rise in recent years.

(Data for 2020 is only available until Oct-20, furthermore Covid-19 may have affected number of conflicts, hence I have selected 2019 for observation)

2) President Rodrigo Duterte’s war on drugs since July 2016.

Philippines: The significant spike in fatalities due to violence against civilians in July-2016 is attributed to President Duterte’s ‘war on drugs’ after taking office in June 2016. This is also prevalent in the top actors involved in these conflicts – anti-drug vigilantes and polices forces of the Philippines. Conflicts happened all over Philippines with majority located on the main island of Luzon (top of Philippines).

3) Protests and Insurgencies in Thailand

Across the past 10 years, Thailand mainly has two types of conflicts namely Protests by Protesters and “Political Violence by the Malay Muslim Separatists.

Thailand’s protests are primarily located in Bangkok and relatively non-violent given the 0 reported fatalities count. However, protests peaked in late 2020 – these pro-democracy protestors seek royal reforms and the Prime Minister Prayuth Chan-ocha’s resignation.

A deeper analysis reveals that these conflicts by the insurgents (Malay Muslim Separatists) are located in the southern part of Thailand, near the border of Malaysia. However, since the month of May-2018, during which a significant number of strategic developments occurred, the number of events has tapered down.

4) The Rohingya Conflict - Myanmar

Myanmar’s conflict is primarily located at the Rakhine state. This conflict is otherwise known as the Rohingya crisis. These events are generally violent with a high fatality toll. The deadliest month was August 2017, with a reported death toll of 1,043 due to Myanmar’s military crackdown after a police post was attacked by Rohingya militants.

5) Indonesia’s conflicts in 2020 were mainly protests/riots and some battles in West Papua

In the year 2020 (data only available until Oct-2020), despite Covid-19, Indonesia experienced the greatest number of events per country within selected South East Asian countries. Events in 2020 were mainly protests, which spiked significantly in Oct-2020 due to a contentious new job creation law, these protests were generally non-violent with 0 death toll.

Whereas battles in West Papua are a result of an ongoing West Papua conflict.

5. Data visualisation steps

5.1 Data Preparation

On the landing page of Tableau, under Connect > To a file > More. And select the data file.

5.2 Data visualisation steps

1) Map

1. Plot map Drag [Longitude] to Columns and [Latitude] to Rows.

For both fields - right click and change attributes to dimension.

2. Drag [Event type] to Marks > Colour, drag [fatalities] to size.

3. Format size of bubble and reduce opacity according to preference, here have used opacity of 60%.

4. Change colour palette of event type according to figure below. I have used the reddish-orange shades for violent events - battles/violence against civilians/explosions and blue shades for non-violent events - riots/protests/strategic developmet.

5. Use dark background for map.

6. Filters i) Add [fatalities] to filter - this allows user to filter events by fatalities range.

ii) Add Country to filter - this allows users to select specific countries

iii) Add Year to filter - map will only display events / fatalities for that year.

7. Add details in tooltip.

8. Change title and update formatting.
   

9. Change bubble size - so that bubble size doesn’t change when filtering by countries as it may skew the sizes to be larger/smaller depending on the total fatalities of the country/select filter.

2) Time series toggle

1. Create parameters

2. Create calculated fields i) True and False

ii) Toggles

3. Create toggle button

4. Select Shape - you will need two icons to represent yes and no separately. Use parameter toggle to change between yes and no, then assign icons.
   For more details on importing shapes, refer to Tableau blog.

5. Edit tooltips to include the following:

6. Format worksheet - remove background colour, change view to “Entire View”.

Action for toggling will be added later when creating the time series chart.

3) Time series chart

1. Create parameters

i) Date level selector. Value should be in lowercase.

ii) Events count or fatalities

2. Create calculated fields i) No of records

ii) Date selector

iii) Events or fatalities

3. Add date range filter - to include data from 2015 to 2020.

4. Drag [date selector] to Columns and [Events or fatalities] to rows.

Update the attributes of date selector accordingly.

5. Drag [Event type] to colour and change chart type to Bar.

6. Rearrange chart by dragging variables to this order.

7. Add [Fatalities toggle] to chart.

Dual and synchronise axis.

Change colour to dark red and increase size.

8. Format chart i) Remove header of date and axis of fatalities.

ii) Remove gridlines, worksheet colour to #333333 and font colour to white.

9. Add toggle action

10. Add tooltip i) Create calculated field

ii) Drag [Events or fatalities text] to tooltip and edit tooltip accordingly:

4) Events legend

1. Drag [Event type] to rows and [Event or fatalities] to Columns

2. Drag [Event type] twice to colour and label.

3. Click on label, change alignment to left, edit text accordingly. Change font colour to white.

4. Sort event type

5. Format chart.

i) Remove x-axis header

ii) Remove gridlines.

5) Countries legend

1. Drag [Event/Fatalities] to Columns and [Country] to Rows.

2. Sort [Country] by [Event/fatalities] in descending order.

3. Labels

4. Tooltip

5. Format i) Remove a-axis header

ii) Remove chart borders

iii) Format lines and shadings

6) Actors

1. Drag [Actor1] to Filters pane. Select fields accordingly to create slider to filter top n actors.

2. Drag [Actor1] to rows and [Events or fatalities] to columns.

3. Click on label, change alignment to left, edit text accordingly. Change font colour to white.

4. Change bar colour to red.

5. Sort chart.

6. Update title, font to white.

7. Remove x-axis header

8. Format worksheet - remove gridlines, change worksheet background to grey-black, change font to white

7) Snapshots - fatalities

1. Drag [Fatalities] to Text.

2. Click on text > set alignment to left > change text to the following:

8) Snapshots - events

1. Drag [No of records] to Text.

2. Click on text > set alignment to left > change text to the following:

9) Notes legend for time series

1. Create calculated field

2. Create icon

3. Edit tooltip accordingly.

4. Format

10) Country by event type

1. Duplicate ‘Country’ tab and drag [Event type] to colour

2. Remove marks labels

3. Arrange event type colour

4. Add headers, high field labels for row and change font colour to white

11) Dashboard

1. Go to “Map” tab. i) Apply filter to year.

ii) Apply filter to fatalities, similar to to [Year] filter use Apply to worksheet > Selected worksheet and apply the filter below.

2. For all tabs, ensure that the view is in “Entire View”

3. Dashboard size - change size to automatic.

4. Arrange layout according to sketch - Drag and drop sheets into frame - map, time series and Actors.

5. Add remaining charts

6. For the filters, change all the floating.

Arrange and format accordingly.

7. Select each of the 3 charts (Top n actor, Time series, map) and change dashboard background to #333333

8. Change background colour

9. Add blank above Top n Actors chart so that title is below map title.

10. Used floating > add text.

11. Add action

12. Add source

13. Add author

14. Add an instructional overlay. i) Select floating & vertical -> drop over dashboard.

This will pop up (see image below) > expand vertical object to all ends of the dashboard.

ii) Select tiled > images > drop over vertical object.

iii) Select image and check following boxes. (Hints: My instructions picture was created using powerpoint.)

iv) Change background colour to white and opacity to 68%

v) Add show/hide button and arrange according to instructions. Change borders and background to white.

Test out your overlay instructions.

16. Add “< Click here for instructions using Floating & Text”, place to the right of instructions button.

17. Add Country by event type overlay Repeat step 15 above for ‘Country by Event’ tab.

For step (ii) instead of selecting “images”, select the sheet. Place chart over ‘Country’ chart

For (iv) change background colour to grey and opacity to 100%.

Cartoon vector created by vectorjuice