Difference between revisions of "HappinessWatch: Proposal v3"

From Visual Analytics for Business Intelligence
Jump to navigation Jump to search
(Created page with "<!-- Back Button--> <big> <--- Back to Project Groups </big> <!--Logo--> center|300px <!-- Header Start --> {|s...")
 
 
(22 intermediate revisions by the same user not shown)
Line 42: Line 42:
  
 
<br/>
 
<br/>
−
After further discussion and consultations with Prof. Kam, our team concluded that our initial problem statement and methodology may not be the best approach to analysing suicide data. There was insufficient evidence to draw correlations between socioeconomic factors and suicide rates, as suicide is a complex issue encompassing many factors that cannot be generalised. As such, we decided to pivot our project's direction slightly. Instead of attempting to make conclusions, we will focus on creating a visualisation that allows for exploration and comparison of suicide rates globally.  
+
Upon exploring the data and consulting Prof. Kam, our team decided to focus on visualising global happiness instead due to data constrains and to prevent making false correlations between suicide rates and happiness scores.  
 +
 
 
<br>
 
<br>
 
__TOC__
 
__TOC__
Line 50: Line 51:
 
<!-- Problem and Motivation -->
 
<!-- Problem and Motivation -->
 
==<div style="background: #ef8822; padding: 15px; font-weight: bold; line-height: 0.3em; text-indent: 0px;font-size:20px"><font face="Arial" color=#fbfcfd> PROBLEM & MOTIVATION </font></div>==
 
==<div style="background: #ef8822; padding: 15px; font-weight: bold; line-height: 0.3em; text-indent: 0px;font-size:20px"><font face="Arial" color=#fbfcfd> PROBLEM & MOTIVATION </font></div>==
−
<p>Death by suicide is an extremely complex issue that causes pain to hundreds of thousands of people every year around the world. Close to 800 000 people die due to suicide globally every year, which is one person every 40 seconds. In Singapore, suicide is the leading cause of death for those aged 10-29.
+
<p>Traditionally, a country’s well-being has been measured on economic variables like GDP or unemployment rate. However, no institution, nation or group of people can really be properly understood without also factoring in a number of other elements. One of these key elements is happiness. What contributes to a country’s happiness? Why are some countries happier than others? Are there any trends or patterns we can discern from the available data? With reference to the World Happiness Report, we attempt to visualize the factors that contribute to a country’s happiness on a global scale.
−
<br/>
 
−
<p>The mortality data from the WHO suggests that the prevalence and characteristics of suicidal behavior vary widely between different communities, in different demographic groups and over time. One important source of heterogeneity, both globally and within countries, is gender: suicide rates are much higher for males, particularly in high-income countries. Interestingly in Singapore, females are more likely to be diagnosed with depression and attempt suicide, but males accounted for more than 71% of all suicides in Singapore in 2018. But are there other similarities or patterns that we can discern from the available data? Often seen as a taboo topic, suicide is a very real problem that is not talked about enough. Through our visualisation, we hope to be able to shed some light on and spur conversations on this issue.
 
 
<br/>
 
<br/>
 +
<br>
 +
  
 
<!-- Objectives -->
 
<!-- Objectives -->
Line 59: Line 60:
 
==<div style="background: #ef8822; padding: 15px; font-weight: bold; line-height: 0.3em; text-indent: 0px;font-size:20px"><font face="Arial" color=#fbfcfd>OBJECTIVES</font></div>==
 
==<div style="background: #ef8822; padding: 15px; font-weight: bold; line-height: 0.3em; text-indent: 0px;font-size:20px"><font face="Arial" color=#fbfcfd>OBJECTIVES</font></div>==
 
<p>In this project, we are hope to create a visualization that enables the following:</p>
 
<p>In this project, we are hope to create a visualization that enables the following:</p>
−
# Identify regions or countries with high/low suicide rates
+
# Identify regions or countries with the highest happiness scores
−
# Visualise the relationship between Happiness Index scores and suicide rates
+
# Visualise the happiness scores over time
−
# Exploration of suicide demographics by country
+
# Explore the factors contributing to happiness score
−
# Comparison of suicide demographics between countries
+
# Comparison of happiness scores and its factors across countries
−
# A more in-depth breakdown of Japan's suicide demographics
 
−
<br>
 
−
As mentioned, the reasons behind suicide is complex and cannot be generalised. Through our visualization we hope to show that generic measures like political stability or GDP per capita do not provide a general explanation for suicide. There are definitely many deep-seated issues or underlying cultural norms that could contribute to suicide rates as well. Ideally a further breakdown of the suicide demographics should be conducted for each country, as suicide rates within different parts of a country could vary as well. However, such data is not readily available. We managed to find detailed data on Japan's suicide rates and thought it would make for a good visualisation of the varying suicide rates within a country. Japan has long been known to have high suicide rates, posited to be an amalgamation of its aging population, overwork and even cultural glorification of suicide ("Seppuku"). As such, we decided to breakdown Japan's suicide figures across different prefectures. 
 
 
<br/>
 
<br/>
−
 
 
<!-- Selected Dataset -->
 
<!-- Selected Dataset -->
  
Line 78: Line 75:
 
! style="font-weight: bold;background: #8c8c8c;color:#fbfcfd;width: 30%" | Data Attributes
 
! style="font-weight: bold;background: #8c8c8c;color:#fbfcfd;width: 30%" | Data Attributes
 
! style="font-weight: bold;background: #8c8c8c;color:#fbfcfd;" | Why this Dataset?
 
! style="font-weight: bold;background: #8c8c8c;color:#fbfcfd;" | Why this Dataset?
−
|-
 
−
| <center>Suicide Rates Overview (1985 to 2016)<br/>
 
−
(https://www.kaggle.com/russellyates88/suicide-rates-overview-1985-to-2016) </center>
 
−
||
 
−
* Suicide rate by country
 
−
* Suicide demography (Age/Gender)
 
−
* Economic data (GDP)
 
−
||
 
−
<center>This dataset will be our main source of global suicide data.
 
 
|-
 
|-
 
| <center>World Happiness index 2019<br/>
 
| <center>World Happiness index 2019<br/>
Line 94: Line 82:
 
* Individual segment scores for each country (Freedom of speech, social support, etc)
 
* Individual segment scores for each country (Freedom of speech, social support, etc)
 
||  
 
||  
−
<center>This dataset will be used for comparison to each country's suicide rates, to visualise the relationship between "Happiness" and suicides. Although it may not be the best measure of happiness, we will make do with it as it is the most comprehensive dataset available for happiness scores worldwide.</center>
+
<center>Happiness cannot be easily measured, but the data from the World Happiness Report is the most consistent option with sufficient data points across the past decade. Hence, this will be our choice of data for analysis.</center>
−
|-
 
−
| <center>Japan Suicide Statistics(Yearly)<br/>
 
−
(https://www.npa.go.jp/publications/statistics/safetylife/jisatsu.html/) <br/> </center>
 
−
||
 
−
* Suicide rate by prefecture
 
−
* Suicide rate breakdown by gender and age group
 
−
* Suicide rate breakdown by profession and reason of suicide
 
−
||
 
−
<center>This dataset will be used for in-depth visual analysis for Japan. Compiled and released by Japan's National Police Agency, it contains a detailed breakdown of the suicide demographics across prefectures in Japan.</center>
 
 
|-
 
|-
 
|}
 
|}
−
 
 
<br/>
 
<br/>
−
 
−
<!-- Related Work -->
 
  
 
==<div style="background: #ef8822; padding: 15px; font-weight: bold; line-height: 0.3em; text-indent: 0px;font-size:20px"><font face="Arial" color=#fbfcfd>RELATED WORKS</font></div>==
 
==<div style="background: #ef8822; padding: 15px; font-weight: bold; line-height: 0.3em; text-indent: 0px;font-size:20px"><font face="Arial" color=#fbfcfd>RELATED WORKS</font></div>==
Line 120: Line 96:
 
|-
 
|-
 
|  
 
|  
−
<p><center>'''An interactive dashboard for worldwide suicide data 1985-2015''' </center></p>
+
<br>
−
[[File:Worldmap suicide rate.png|500px|center]]
+
<p><center>'''Choropleth Map by Region''' </center></p>
−
<p><center>'''Source''': https://www.kaggle.com/tavoosi/suicide-data-full-interactive-dashboard/#data</center></p>
+
[[File:HW_related1.png|500px|center]]
 +
<br>
 +
<p><center>'''Comparison Table of Two Countries' Happiness Score Components''' </center></p>
 +
[[File:HW_related2.png|500px|center]]
 +
<br>
 +
<p><center>'''Source''': https://countryeconomy.com/demography/world-happiness-index</center></p>
 +
<br>
 
  ||  
 
  ||  
−
* This dashboard used a combination of a choropleth map and bar chart, which aids in recognition of regions with high/low suicide rates and identifying the rank of a particular country.
+
* By visualising the happiness rankings, it enables the viewer to make easy comparison between continents.
−
* The coordinated color scale makes it easy for users to understand data from both charts.
+
* The comparison table however, makes it difficult to compare the differences in components at a glance.
 
|-
 
|-
 
|  
 
|  
−
<p><center>'''Animated time-series bar chart''' </center></p>
+
<br>
−
[[File:SW_related2.png|500px|center]]
+
<p><center>'''Stacked Bar Chart of Individual Happiness Score Components''' </center></p>
−
<p><center>'''Source''': https://ourworldindata.org/suicide</center></p>
+
[[File:HW_related3.png|500px|center]]
 +
<br>
 +
<p><center>'''Bar Chart Measuring Change in Happiness Score Over the Years''' </center></p>
 +
[[File:HW_related4.png|500px|center]]
 +
<br>
 +
<p><center>'''Source''': https://worldhappiness.report/ed/2019/</center></p>
 +
<br>
 
  ||  
 
  ||  
−
* Visualises the changes in countries with the highest suicide rate over the years
+
* The sorted stacked bar chart visualises the component breakdown of each countries' happiness score. However, it is difficult to compare the differences in individual components across countries.
−
* Allows user to select specific countries that they wish to include in the comparison
+
* Visualising the change in happiness scores with a bar chart makes the changes in individual country's score obvious, but there is no indication of the actual score. (e.g. low to high? Or high and higher?)
 
|-
 
|-
 
|  
 
|  
−
<p><center>'''Animated time-lapse of suicide rates between developed and developing countries''' </center></p>
+
<br>
−
[[File:SW_related3.png|500px|center]]
+
<p><center>'''Interactive Radial Stacked Bar Chart of Global Happiness Score Components''' </center></p>
−
<p><center>'''Source''': https://medium.com/@garytse_91587/world-suicide-rates-a-visualization-636c6f2f1e15</center></p>
+
[[File:HW_related5.png|500px|center]]
 +
<br>
 +
<p><center>'''Source''': http://www.benscott.co.uk/wdvp/</center></p>
 +
<br>
 
  ||  
 
  ||  
−
* The visualization allows users to filter the data based on the type of countries(developed/developing) over time.
+
* The interactivity makes for an enjoyable user experience. The use of filtering enables users to explore the individual components as well, for more in depth analysis.
−
* By breaking down the data the user can target specific countries of interest without getting overwhelmed by the large amount of data.
+
* The radial layout causes some difficulty in looking for a specific country.
 
|}
 
|}
 
<br/>
 
<br/>
Line 157: Line 148:
 
|-
 
|-
 
|  
 
|  
−
<p><center>'''The New Zealand Labour Market Dashboard''' </center></p>
+
<p><center>'''Scatterplot Quadrant Analysis''' </center></p>
−
[[File:Two-point-line-graph.png|500px|center]]
+
[[File:HW_inspiration1.png|500px|center]]
−
<p><center>'''Source''': https://mbienz.shinyapps.io/labour-market-dashboard_prod/</center></p>
+
<p><center>'''Source''': http://www.analyticshero.com/2012/09/11/how-to-use-scatterplot-quadrant-analysis-with-your-web-analytics-data/</center></p>
 
  ||  
 
  ||  
−
* Effectively visualises value changes between any two years
+
* Effectively visualises value changes between the two axis values
−
* Any increase/decrease is immediately apparent
+
* Overall picture and categorisation by quadrants
−
* However, fluctuations between the selected points cannot be visualised.
 
 
|-
 
|-
 
|  
 
|  
Line 179: Line 169:
 
==<div style="background: #ef8822; padding: 15px; font-weight: bold; line-height: 0.3em; text-indent: 0px;font-size:20px"><font face="Arial" color=#fbfcfd>PROPOSED STORYBOARD</font></div>==
 
==<div style="background: #ef8822; padding: 15px; font-weight: bold; line-height: 0.3em; text-indent: 0px;font-size:20px"><font face="Arial" color=#fbfcfd>PROPOSED STORYBOARD</font></div>==
  
−
Our proposed application will consist of four pages:
+
Our proposed application will consist of three pages:
−
 
 
−
=== LANDING PAGE ===
 
−
<!-- Table -->
 
−
{| class="wikitable" style="background-color:#FFFFFF;" width="100%"
 
−
|-
 
−
! style="font-weight: bold;background: #8c8c8c;color:#fbfcfd;width: 50%;" | Proposed Layout
 
−
! style="font-weight: bold;background: #8c8c8c;color:#fbfcfd;" | Description
 
−
|-
 
−
|
 
−
[[File:SuicideWatch_v2_home.jpg|700px|center]]
 
−
||
 
−
This page will serve as an introduction to our problem and objectives, to give the viewer an overview of our project. The viewer can navigate to specific sections by clicking on the respective headers.
 
−
|}
 
−
<br/>
 
  
 
=== OVERVIEW ===
 
=== OVERVIEW ===
Line 204: Line 180:
 
|-
 
|-
 
|  
 
|  
−
[[File:SuicideWatch_v2_overview.jpg|700px|center]]
+
[[File:HW_Overview.jpg|700px|center]]
 
  ||  
 
  ||  
  
−
# '''Bar Chart and Choropleth Map'''
+
# '''Sorted Stacked Bar'''
−
#* The bar chart will show the suicide rates globally, sorted in ascending or descending order.
+
#* The Stacked Bar shows the component breakdown of each country's Happiness Index Score.
−
#* Both charts will be coloured with the same color intensity scale, to allow for ease of comparison.
+
#* Users can choose to sort the bar by the overall score or individual components
−
#* When selecting a particular country on either maps, the corresponding object on the other chart will be highlighted as well.
+
# '''Choropleth Map'''
 +
#* The Choropleth fills in the world map with varying color intensities based on the Happiness Index Scores.
 +
#* Enables identification of "happy" or "unhappy" clusters/regions.
 
# '''Scatterplot'''
 
# '''Scatterplot'''
−
#* The scatterplot of suicide rates vs happiness score visualises the relationship between these 2 measures.  
+
#* Plots the Happiness Index Score changes across two selected years.  
−
#* Depending on their positions within the 4 quadrants, we can get interesting insights on a specific country's suicide numbers. For example, a country that ranks highly on the happiness index but has high suicide rates could represent the presence of some underlying issues that lead to increased suicides.
+
#* Upper left and lower right quadrants represent decreases and increases in Happiness Index Scores respectively.
 +
# '''Ridge'''
 +
#* Visualises the distribution of components contributing to the Happiness Index Scores.
 
|}
 
|}
 
<br/>
 
<br/>
  
−
=== BREAKDOWN ===
+
=== COUNTRY COMPARISON ===
 
<!-- Table -->
 
<!-- Table -->
 
{| class="wikitable" style="background-color:#FFFFFF;" width="100%"
 
{| class="wikitable" style="background-color:#FFFFFF;" width="100%"
Line 225: Line 205:
 
|-
 
|-
 
|  
 
|  
−
[[File:SuicideWatch_v2_demographic.jpg|700px|center]]
+
[[File:HW_Comparison.jpg|700px|center]]
 
  ||  
 
  ||  
−
# '''Summary Card'''
+
# '''Radar Chart:''' Component breakdown of the Happiness Index Scores of each selected country.
−
#* A combination of statistics and charts that summarises the demographics of a particular country's suicides.
+
# '''Ridge Plot:''' Compares the distribution of components contributing to the Happiness Index Score for each selected country.
−
#* '''Radar Chart:''' Component breakdown of the Happiness Index.  
+
 
−
#* '''Line Chart:''' Time series of the country's suicide figures, separated by age and gender.
 
−
# '''Country Comparison'''
 
−
#* Viewers can add another Summary Card for side-by-side comparison between 2 or more countries' suicide demographics.
 
 
|}
 
|}
 
<br/>
 
<br/>
  
−
=== CASE STUDY: JAPAN ===
+
=== COUNTRY TIME-SERIES ===
−
To show that country level analysis may still be too generalised, this page is a further breakdown of Japan's suicide statistics at the prefecture level.
 
 
<!-- Table -->
 
<!-- Table -->
 
{| class="wikitable" style="background-color:#FFFFFF;" width="100%"
 
{| class="wikitable" style="background-color:#FFFFFF;" width="100%"
Line 245: Line 221:
 
|-
 
|-
 
|  
 
|  
−
[[File:SuicideWatch_v2_japan.jpg|700px|center]]
+
[[File:HW_CountryTS.jpg|700px|center]]
 
  ||  
 
  ||  
−
# '''Cartogram of Japan'''
+
# '''Comparison Line:''' Compares the selected country's Happiness Index Scores over time to the average score.
−
#* An animated cartogram of Japan that shows the suicide rate of each prefecture.
+
# '''Component Breakdown Line:''' Plots the component scores for the selected country over time.
−
#* Users can select 1 or more prefectures for comparison across the other charts.
 
−
# '''Reasons for Suicide and Occupations'''
 
−
#* '''Radar Chart:''' A breakdown of the reasons for suicide in the prefecture.
 
−
#* '''Stacked Bar Chart:''' A breakdown of the occupations of the people who committed suicide in the prefecture.
 
−
#* The radar and bar chart will layer themselves according to the prefectures selected to allow for easy comparison.
 
−
# '''Connected Dot Plot'''
 
−
#* Visualises the changes in suicide rates over time, based on prefectures selected.
 
 
|}
 
|}
 
<br/>
 
<br/>
Line 304: Line 273:
  
 
==<div style="background: #ef8822; padding: 15px; font-weight: bold; line-height: 0.3em; text-indent: 0px;font-size:20px"><font face="Arial" color=#fbfcfd>REFERENCES</font></div>==
 
==<div style="background: #ef8822; padding: 15px; font-weight: bold; line-height: 0.3em; text-indent: 0px;font-size:20px"><font face="Arial" color=#fbfcfd>REFERENCES</font></div>==
−
* WHO - World Suicides (https://www.who.int/news-room/fact-sheets/detail/suicide)
+
* Semantic Dashboard (https://github.com/Appsilon/semantic.dashboard)
 
* R Shiny Gallery (http://shiny.rstudio.com/gallery/)
 
* R Shiny Gallery (http://shiny.rstudio.com/gallery/)
 
* World Happiness Report (https://worldhappiness.report)
 
* World Happiness Report (https://worldhappiness.report)

Latest revision as of 12:31, 19 November 2019

<--- Back to Project Groups

Team 3 - HappinessWatch Logo.png

ABOUT US

PROPOSAL

POSTER

APPLICATION

RESEARCH PAPER


Version 3


Upon exploring the data and consulting Prof. Kam, our team decided to focus on visualising global happiness instead due to data constrains and to prevent making false correlations between suicide rates and happiness scores.



PROBLEM & MOTIVATION

Traditionally, a country’s well-being has been measured on economic variables like GDP or unemployment rate. However, no institution, nation or group of people can really be properly understood without also factoring in a number of other elements. One of these key elements is happiness. What contributes to a country’s happiness? Why are some countries happier than others? Are there any trends or patterns we can discern from the available data? With reference to the World Happiness Report, we attempt to visualize the factors that contribute to a country’s happiness on a global scale.

OBJECTIVES

In this project, we are hope to create a visualization that enables the following:

  1. Identify regions or countries with the highest happiness scores
  2. Visualise the happiness scores over time
  3. Explore the factors contributing to happiness score
  4. Comparison of happiness scores and its factors across countries


SELECTED DATASETS

Dataset/Source Data Attributes Why this Dataset?
World Happiness index 2019
(https://worldhappiness.report/ed/2019/)
  • Overall Happiness Rankings of Countries Worldwide
  • Individual segment scores for each country (Freedom of speech, social support, etc)
Happiness cannot be easily measured, but the data from the World Happiness Report is the most consistent option with sufficient data points across the past decade. Hence, this will be our choice of data for analysis.


RELATED WORKS

Example Takeaways


Choropleth Map by Region

HW related1.png


Comparison Table of Two Countries' Happiness Score Components

HW related2.png


Source: https://countryeconomy.com/demography/world-happiness-index
  • By visualising the happiness rankings, it enables the viewer to make easy comparison between continents.
  • The comparison table however, makes it difficult to compare the differences in components at a glance.


Stacked Bar Chart of Individual Happiness Score Components

HW related3.png


Bar Chart Measuring Change in Happiness Score Over the Years

HW related4.png


Source: https://worldhappiness.report/ed/2019/
  • The sorted stacked bar chart visualises the component breakdown of each countries' happiness score. However, it is difficult to compare the differences in individual components across countries.
  • Visualising the change in happiness scores with a bar chart makes the changes in individual country's score obvious, but there is no indication of the actual score. (e.g. low to high? Or high and higher?)


Interactive Radial Stacked Bar Chart of Global Happiness Score Components

HW related5.png


Source: http://www.benscott.co.uk/wdvp/
  • The interactivity makes for an enjoyable user experience. The use of filtering enables users to explore the individual components as well, for more in depth analysis.
  • The radial layout causes some difficulty in looking for a specific country.



DESIGN INSPIRATIONS

Example Takeaways

Scatterplot Quadrant Analysis

HW inspiration1.png

Source: http://www.analyticshero.com/2012/09/11/how-to-use-scatterplot-quadrant-analysis-with-your-web-analytics-data/
  • Effectively visualises value changes between the two axis values
  • Overall picture and categorisation by quadrants

NBA Player Statistics Visualization

SW Inspiration2.png

Source: https://wilsoncernwq.github.io/NBAstatsVIS/documents/Proposal.pdf
  • The concept of a "Summary Card" makes it easy to compare two players when put side by side.
  • The use of radar chart can effectively break down a measure with multiple components (e.g. Happiness Index)



PROPOSED STORYBOARD

Our proposed application will consist of three pages:

OVERVIEW

This page will provide the viewer with an overview of global suicide rates and overall happiness scores.

Proposed Layout Description
HW Overview.jpg
  1. Sorted Stacked Bar
    • The Stacked Bar shows the component breakdown of each country's Happiness Index Score.
    • Users can choose to sort the bar by the overall score or individual components
  2. Choropleth Map
    • The Choropleth fills in the world map with varying color intensities based on the Happiness Index Scores.
    • Enables identification of "happy" or "unhappy" clusters/regions.
  3. Scatterplot
    • Plots the Happiness Index Score changes across two selected years.
    • Upper left and lower right quadrants represent decreases and increases in Happiness Index Scores respectively.
  4. Ridge
    • Visualises the distribution of components contributing to the Happiness Index Scores.


COUNTRY COMPARISON

Proposed Layout Description
HW Comparison.jpg
  1. Radar Chart: Component breakdown of the Happiness Index Scores of each selected country.
  2. Ridge Plot: Compares the distribution of components contributing to the Happiness Index Score for each selected country.


COUNTRY TIME-SERIES

Proposed Layout Description
HW CountryTS.jpg
  1. Comparison Line: Compares the selected country's Happiness Index Scores over time to the average score.
  2. Component Breakdown Line: Plots the component scores for the selected country over time.


PROJECT TIMELINE

Team 3 - SuicideWatch Timeline.png
Team 3 - SuicideWatch Gantt.png


KEY CHALLENGES

Challenge Mitigation

Inexperienced with Creating and Designing Visualisations

  • Engage in hands-on practice during and after class.

Inexperienced with R and R Shiny

  • Make full use of the DataCamp resources provided
  • Self-directed and peer learning

Limited Access to Sensitive Suicide Data

  • Acquire data from various sources and conduct data cleaning to organize the data.
  • Make do with what we can get, down scope the project accordingly.

Time and Workload Constrains

  • The team will come up with a reasonable project timeline based on everyone's ability and capacity.
  • Set milestones and adjust the timeline accordingly based on the team's progress.


REFERENCES


COMMENTS

Feel free to leave us some comments or feedback!

Name Comment/Feedback

Your Name

  • Comment

Your Name

  • Comment
Retrieved from ‘https://wiki.smu.edu.sg/1920t1is428g1/index.php?title=HappinessWatch:_Proposal_v3&oldid=7589’