Introduction

Suicide is a painful and complex social phenomenon due to which the lifespan of so many people gets shortened before time. The World Health Organization (WHO) and the Global Burden of Disease study estimate that almost 800,000 people die from suicide every year. That’s one person every 40 seconds. (Ritchie, 2015)

It is the 12th leading cause of death in the US and for each person that commits suicide, there are twenty-five more who have attempted it. (Suicide Statistics and Facts, n.d.)  Suicide is more common than homicide across most countries in the world – often as much as ten to twenty times higher. (Ritchie, 2015)The factors which put an individual at a higher risk of committing suicide are mental health disorders (especially depression and alcohol use disorder), substance abuse, experiencing conflict, disaster, violence, abuse or a loss and sense of isolation, negative life events, and relationship breakups. The suicide rates are also the highest amongst vulnerable groups like refugees and migrants; indigenous peoples; lesbian, gay, bisexual, transgender, intersex (LGBTI) persons; and prisoners. However, the greatest risk factor is a prior suicide attempt.

Suicide is a global phenomenon but more than 77% of the global suicides that took place in 2019 were in low and middle-income countries. (WHO, 2021) Ingestion of pesticides, hanging, and firearms are among the most common methods of suicide globally. In terms of viewing it through the lens of gender, Suicide is more common in men than women in all countries (Ritchie, 2015) 

The above suicide statistics as painful as they are on their own, fail to take into account the countless other lives of the family, friends, and loved ones that get destroyed with each person who ends their life. This project aims to identify the most vulnerable groups by using the data collected on suicides worldwide and increase awareness about the factors which put individuals at high risk. The research questions that the visualizations seek to answer are:

1. Which countries have higher suicide rates?
2. How do the suicide rates differ across sex?
3. Which age groups are most vulnerable?
4. How does the prevalence of depression affect the suicide rate across genders?

Inspiration

This project was inspired by various kinds of visualizations. I started by studying the visualizations by w.e.b Dubois where he represented the invisible lives of the African American community in the Paris exposition in 1900. I was struck by the simplicity by which complex topics and demographic factors were represented. Another, source of inspiration was the article in Our world in data which studies the phenomenon of suicide very extensively. It broke down the topic and related it to age, gender, countries, and mental health disorders. 

Process

1. Collecting data

In order to create my visualizations I used five different datasets from the following sources: Why are suicide rates higher for men, Prevalence of mental disorders, Age standardized suicide rates, and WHO suicide data. All of these data sources were found on Kaggle and before importing the data I had to consolidate a few of them in Excel. 

2. Importing the data to Tableau

 I decided to use Tableau which is a comprehensive data visualization tool, in order to represent my data. Since I had five different datasets the first step was to connect them by creating relationships on the basis of common attributes like countries or years.

3. Planning the dashboard structure

I decided to create an interactive dashboard that would allow my viewers the greatest flexibility to interact with my visualizations. I divided my dashboard of suicide data into three areas of focus (Geography, Gender, and Mental health). 

4. Creating individual visualizations

Once I had my data in place, I started by building individual visualizations. In each visualization, I highlighted gender, age, and the country as they were very important to the story that I was telling. I used choropleth maps, bar charts, and line charts as my primary techniques.

5. Designing the dashboard

After I had my individual visualizations ready I decided to add the colors and styling to it. I wanted to keep the theme dark as this was a serious topic and wanted to use the color red to denote the loss of lives due to suicide. I used pink and blue for denoting gender comparisons. I wanted to give my users the ability to interact with my data so wherever possible I added exposed filters for them. After I styled all my sheets I created the first draft of the dashboard which I would use for user testing.

Usability testing

The user research technique that I used was moderated remote user testing with two participants. I defined four tasks and four high-level questions. I was able to find ten usability issues based on the testing methodology.

Research Goal

My research goal was to find out how intuitive and clear my dashboard was about showing which groups were most vulnerable to suicide prevalence and if my participants could filter the data being shown to them.

Recruiting participants

I recruited two male participants for conducting the usability testing. The reason for that was that I wanted to ask them questions about their emotional support network in order to understand who men rely on for advice and aid during difficult situations and how openly they share their problems. 

Methodology 

I conducted moderated remote user testing with both the participants using zoom. These testing sessions were structured where I had defined my tasks and overall questions and I had shared the link to the Tableau dashboard for them to interact with. I observed their screen and asked them to think aloud while accomplishing the tasks.

Insights

On the basis of the usability test I was able to identify, ten usability issues. Here is the feedback of both the participants and the areas that they were confused with while accomplishing the tasks.

Incorporating feedback

After conducting the Usability testing I understood that the main problem area in my visualizations was a lack of clarity about what my data was representing. I needed to make my titles clearer and add supporting text with each visualization. I also needed to label my charts and graphs with the exact data that it was representing. 

I realized that the color in the first part of my visualization was confusing as a scale of red colors was being used in my map to show the suicide rate in countries. When the same color pattern was used, users automatically assumed that the light-colored red bar represented less suicide rates even if the bar length was higher. I decided to keep the same red color for the five highest countries in terms of suicide rate so that the bar length was the only differentiator.

Another change done by me was making my filters more intuitive. Instead of having checkboxes for years which seemed confusing, I added a slider. Also for selecting the genders I made it a radio button to allow for only a single selection.

Lastly, I looked at the overall structure of my dashboard and added more negative spacing between various topics to make my groupings clearer, and kept the filters for each visualization right next to it.

Here is the design for my final dashboard after the usability testing process.

Rationale

I have created an extensive dashboard that shows the relationship between various factors like age, gender, country, and suicide so that I can use data to show which groups are at the highest risk. I wanted to cover this topic as this topic is often brushed under the carpet. In terms of my design choices, I have used a variety of simple but effective tools for data visualizations like choropleth maps, bars, and line charts. I have added supplemental information for each individual graph to make it clearer to understand. 

In terms of the visual design, the goal was to convey the seriousness of the topic and use minimal and effective colors against a dark background to have a lot of contrast and easy legibility. Post my user testing, I have incorporated the insights and feedback on my dashboard to make it more intuitive, especially with the filters used.

Findings

1. Global overview of the suicide rates

The first section of my dashboard focuses on suicide rates globally. In order to create these visualizations, I used the age-standardized suicide rates for 2015. Upon analyzing the map and bar charts, it was clear that the suicide rates were spread globally and the countries at high risk were: Lithuania, Kazakhstan, Ukraine, Uruguay, Russia and Latvia. Based on my usability testing, I added clearer titles and labels which mentioned what exactly was being measured. Another change was making the gender filter select a single value instead of multiple selection checkboxes. I have also added the gender filter for both of these visualizations so that the viewers can see how the global distribution changes.

2. Suicide rates across the genders

The second segment of my dashboard focuses on the suicide rate across genders. I used the Suicide rate per 100k population for creating these visualizations. The suicide rates across the two sexes were very contrasting. The suicide rate for men was consistently greater than for women. As shown in the line chart, the suicide rate for men in 1995 was 3.66 times higher. The same pattern followed through for all age groups and it was the highest in the older age groups in both the genders. Based on my user testing, I made my labels clearer and removed the highlighting tooltip which was obstructing the user’s view.

3. Global overview of depression and suicide rates with depression

The third segment of my dashboard looked at the global prevalence of depression. The reason I decided to incorporate this was that depression is the leading cause of disability worldwide and is a high-risk factor for suicide prevalence. Globally depression rate was the highest in Greenland, I added a slider for selecting years here to track the depression rates across years. Before the user testing, the years were shown in the form of multiple selection checkboxes. The participants found that very confusing so I changed it in my final dashboard. The last visualization shows the suicide rates for people who have depression. Research shows that depression is more prevalent in women. However, the suicide rate for women is much lesser in comparison to men across multiple years.

Recommendation

I really enjoyed working on this project. What was different in this final project was the process of usability testing and I found that very helpful in understanding any problems which can miss our eyes as the designers. I also enjoyed the ability to dive deep into one topic and create multiple forms of visualizations and analyze my topic more extensively. A limitation that I faced in this project was that there wasn’t a single dataset with all the required data points, so there were situations where the data wasn’t as consistent as I would have liked. I also would have liked it if the data for suicide rate was not limited to only the two sexes and took into account the suicide rates for the transgender community for instance and was inclusive. Even with the data limitations, I learned a lot in this project and tried to show my data in the most detailed way possible.

References

Ritchie, H. (2015, June 15). Suicide. Our World in Data. https://ourworldindata.org/suicide

Suicide Statistics and Facts. (n.d.). SAVE. https://save.org/about-suicide/suicide-statistics/

WHO. (2021, June 17). Suicide. https://www.who.int/news-room/fact-sheets/detail/suicide