Author: Patnaik Bikram

Bikram Patnaik

Visualization Link: Fertility vs Life Expectancy

Maternal mortality is much higher in developing than in developed countries- Mahler, 1987

“The world is still ‘we’ and ‘them.’ And ‘we’ here refers to the Western world and ‘them’ is Third World. And what do you mean by Western world? Well, that’s long life and small family, and Third World is short life and large family.” – This is a podcast conversation that I heard last week playing in my friend’s car.

Well, the importance of quantifying the loss of life caused by maternal mortality in a population is widely recognized. In 2000, the UN Millennium Declaration identified the improvement of maternal health as one of eight fundamental goals for furthering human development.

Going by the old school definition, maternal mortality ratio (MM Ratio) is obtained by dividing the number of maternal deaths in a population by the number of live births occurring in the same time interval. It depicts the risk of maternal death relative to the frequency of childbearing.

With the help of this visualization we will discuss if we can justify our main claim with proper evidence or is it just a regular misconception which should only interest health experts and statistician. The visualization reviews data from 150 countries and compares Life expectancy (maternal mortality rate) vs maternal fertility for the past 2 centuries. The vertical axis shows the average maternal life span in each country ranging from 10 – 80 years, where high up= longer life, to the bottom= shorter life. The horizontal axis shows the total fertility rate that ranges from 2- 9 children per woman. It’s interesting to see the usage of bubble chart for this, which is primarily used when you represent data that has three or more data series (In this case Life expectancy, total fertility and size of the population) and each containing a set of values.

UNDERSTANDING THE DATA:

Here each country in the world is a bubble,the size of the bubbles represent the population size and color represents regions of the world (see on top right side).Before rolling the years, we notice that in 1800 all the countries had a life expectancy less than 45 years and the children per woman ranged from 4-8 (on an average- 6 children) in a family. For further discussions let’s understand that high mortality rate means less life expectancy and vise-versa. Developed countries like France had a low fertility rate( 4.4- which means they had smaller families) and a life span of 34 yrs where as the developing county, Iran with a smaller population had a high fertility rate (7.1- meaning larger family) and the life span barely crossing 25 years. We see there is a significant difference between the developed and developing countries in terms of both mortality and fertility rate pertaining to factors like lack of adequate medical care, the greater prevalence of infectious diseases and higher incidence of pregnancy. This might seem to be a promising warrant that we want to verify our claim.

BUT!! let’s see what happens when years pass by. Till the 1940s there was no significant difference between the countries on the visualization. Only after 1950, there is a change that is noticeable. China starts moving with better health and improves steadily. All Latin American countries start to move towards smaller families. The blue ones are the Arabic countries,and they have longer life, but no larger families.In the ’80s, Bangladesh still remains similar to the African countries. But then the Bangladeshi imams start to promote family planning and pull the country higher up the life expectancy ladder. And in the ’90s, there was a terrible HIV epidemic that takes down the life expectancy of the African countries and all the rest of the countries move up where there are longer lives and smaller families, and we have a completely new world.

Let me make a comparison directly between the United States of America and Vietnam. In 1964, America had small families and long life; Vietnam had large families and short lives. The data during the war indicate that even with all the death, there was an improvement of life expectancy. By the end of the year, the family planning started in Vietnam; they went for smaller families. While United States had longer life, keeping family size. In the ’80s, Vietnam gives up Communist planning and goes for market economy, and it moves faster even than social life. And today in Vietnam, we have  the same life expectancy and the same family size as in United States, 1974, by the end of the war. With this all the previous warrants fails and now this comparison acts as a strong rebuttal against our original claim.

DRAWBACKS:

Undoubtedly the visualization is amazing in itself, but there are few snags which can alter the statistics if taken into consideration. First, the data collected are only with respect to inter-countries. But what it doesn’t include is the scope to look at the differences among the maternal mortality within the regions of a given country which would give different insights. Second, there is no mention of various age group of women who face higher mortality rate; so that those age group can be targeted for special medical care during pregnancy. Third, there is no comparison with the actual vs target MM ratio data for any given country. There is a high probability that when these factors are combined together it might give us a different picture altogether.

FROM A CRITIQUE’S VIEWPOINT: 

The number of different parameters presented on the interactive dashboard are overwhelming and seems far less from being user-friendly. For a new user it becomes hard and confusing, instead a simple drop down could be introduced to give the audience the flexibility to play around with their desired set of parameters.The simpler it is, the easier it becomes. Second, the visualization doesn’t allow us to see the change in MM Ratio across various countries as year passes. This is very crucial for any government health organisations to plan ahead of time. Third, though we see a bubble comparison between inter-countries but the total mortality figures/data over the years for a particular country is not seen ( neither it’s total comparison with the World data).

ALTERNATIVE APPROACH/MODIFICATION:

Though it’s visually appealing there are certain hiccups with this bubble charts visualization as well. It can be further enhanced and made simpler by adopting certain techniques.

1. The bubble chart earlier didn’t give us a change in MMR for different countries. So, as a modification I would recommend to use 2-D clustered bar graph to show the deviation of MMR across the various countries over a period of time. This is a simpler alternative to visually represent a change in data, as you can see this in below visualization.

2. In this following modification, we make changes to the graph as pointed out in drawback section. Using a simple line chart  we try to contrast two different projections for a given country. In the below visualization, we see the target MMR of Mexico and a comparison of it’s actual stats.

3. Further, we discussed that a bubble comparison between inter-countries is seen in the visualization but there is no individual mortality figures/data for a particular country over the years ( neither it’s total comparison with the World data).. The modified version looks like the below visualization.

4.The following visualization segregates the maternal mortality into different age groups and makes it easier to understand the impact over the spectrum.

CONCLUSION:

We could clearly see from the comparison between countries that in the modern era, all the developed as well as developing countries have the same mortality and fertility rate. Studies suggest that 45% of the potential number of maternal lives saved in developing countries is attributable to fertility decline and 55% of the potential number of maternal lives saved are because of safe motherhood initiatives.

If we don’t look in the data, I think we all underestimate the tremendous change in Asia, which was in social change before we saw the economical change.So rather than over simplifying the fact that only maternal mortality is higher in the ‘Them’ countries as compared to ‘We’ countries, we should try to embrace the changes happening around the world.

References : Princeton, NCBI , Encyclopedia Iranica, Quora, WHO, Mamaye, Wikipedia

Bikram Patnaik

Visualization Link: HIV: Forgotten Epidemic

More PEOPLE DIED OF AIDS IN AFRICA THAN IN ALL WARS ON THE CONTINENT‘ –UN Secretary General, Kofi Annan

Yes you heard it right! As a matter of fact Sub-Saharan Africa carries a disproportionate burden of HIV, accounting for more than 70% of the global burden of infection.

Well, if you grew up in the 1990s, you practically absorbed a degree in AIDS studies just by existing—or at least that’s what it felt like. The years since then have brought better tests and treatments, and we now know more about the virus, but that information isn’t common knowledge. HIV and AIDS has still not fallen off the radar and continues to impact the lives of people in various corners of the globe.

With the help of this visualization we will discuss if we can justify our main claim with proper evidence or is it just another eye-catching headline story.The visualization which we are about to discuss reviews data from 85 countries and compares percentage of adults vs wealth for the past 4 decades. The vertical axis shows the percentage of adults infected with HIV virus ranging from age 15-49 years. The horizontal axis shows the average income per person (GDP per capita) expressed in dollars per person per year. It’s interesting to see the usage of bubble chart for this, which is primarily used when you represent data that has three or more data series (In this case income, % of HIV infected people and size of the population) and each containing a set of values.

UNDERSTANDING THE DATA:

Let’s dive deeper into the viz by understanding it’s working dynamics. On analyzing we see that each country in the world is a bubble,the size of the bubbles represent how many people are infected in a particular country and color represents regions of the world (see on top right side).We start the HIV epidemic cycle and notice that in year 1985,all the countries had an infected population percentage less than 1% but the income ranges broadly from $400 to nearly $40,000. The United States being the richest country had a very little percentage of people infected but the size of the bubble is significant as compared to rest of the countries suggesting that it had the large section of citizens who were infected with this deadly virus. With 5% infection rate Zambia and Uganda were the highly infected countries but with a lower income.

As years pass by, it’s shocking to notice that only the African countries experience a sky rocketing growth rate (highest being 26%) of HIV infection while it stayed low for the rest of the world. As a result of economic slowdown , Africans themselves had neither the resources nor the money to discover vaccines that prevent AIDS, which acts as a strong warrant to our claim.The backing for this warrant can be read in the form of this article. For the past 3 years we have reached a steady state of HIV epidemic. Steady state doesn’t mean that things are getting better, it has just stopped getting worse. Only 1% of the world adult population those who are infected by HIV fall under this steady state, roughly around 40 million (for comparison it equals California population today).

As our main claim revolves around the African sub-continent, we will focus more on them. Let’s take Botswana as our specimen and analyze it. Having an economy better than it’s counterpart African countries, it started low on infection rate but picked up in 2003 before finally declining slowly. Because of better economy Botswana is able to treat people. Those who are treated don’t die of AIDS rather they survive longer and as a result the % wouldn’t come down. Poor African countries like Somalia have lower infection rate because people can’t afford to expensive medical care and die as a result, infact it’s % figures matches with rest of the world. This fact certainly acts as a rebuttal to our claim that all of Africa acts as an incubator for HIV infected people.

DRAWBACKS:

Undoubtedly the visualization is stunning, but there are few snags which can alter the statistics if taken into consideration. First, the data collected are only with respect to inter-countries. But what it doesn’t include is the scope to look at the differences among the HIV infected population within the regions of a given country which would give insights to it’s degree of severity. Second, here we are only talking about the infected % of adults but their is no comparison with the death/ mortality rate. So there is a high probability when combined together it might give us a different picture altogether.

FROM A CRITIQUE’S VIEWPOINT: 

The number of different parameters presented on the interactive dashboard are overwhelming and seems far less from being user-friendly. For a new user it becomes hard and confusing, instead a simple drop down could be introduced to give the audience the flexibility to play around with their desired set of parameters.The simpler it is, the easier it becomes. Also the age group mentioned here ranges from 15-50 years, which fails to segregate the individual age group being infected. The vertical lines on the chart needs to be even spaced. Also, while toggling on the ‘Map’ tab, it gives us an elliptical view of the globe and the bubble of each country fail to sync with their respective geographical location. This can be eliminated by displaying a flat world map view with bubbles corresponding to accurate geographical locations. An additional feature of forecast can be introduced to visualize the future trends and the possible repercussions of this epidemic.

ALTERNATIVE APPROACH/MODIFICATION:

Though it’s visually appealing there are certain hiccups with this bubble charts visualization as well. It can be further enhanced and made simpler by adopting certain techniques.

1. The bubble chart earlier gave us an entire range of age groups. So, as an modification I would recommend to use trend lines to show the % of individual age groups infected. This is a simpler alternative to the control group arrangements, as you can see this in below visualization.

2. In this following modification we can see that all the bubbles sync accurately with their geographical points. As a result we can visually identify that a majority of the bubbles are concentrated in Africa.

3. Further, as we discussed in the drawbacks that the visualization doesn’t show the differences within a country, It can be modified by introducing an additional feature in which by selecting a country say Africa, it will give you an overview of all the data values for the 54 states along with an appropriate color contrast. The modified version looks like the below visualization.

CONCLUSION:

We could clearly see that more than 50% of the African population (0.6 billion) are affected by AIDS and by statistical data around 0.25 million Africans have died due to wars. Thus, it strongly affirms the claim made by the UN secretary general, Kofi Annan.  But at the same time, over simplifying the fact that only African sub-continents are affected by HIV would be a wrong judgement. UNAIDS has provided sufficient data proving that all parts of the world are in the grip of this epidemic virus. So, instead of worrying about the expensive treatments, as a socially responsible person we should focus more on the prevention rather than it’s after effects. As prevention is the only way we can make the world a safer and a better place.

References : Globalissues.org, Discovermagazine.com , Thegaurdians.com, afro.who

Bikram Patnaik

Visualization Link: Money buys Life

‘THE RICH LIVE LONGER EVERYWHERE, BUT FOR THE POOR GEOGRAPHY MATTERS’

Did it ever strike to you that the place our ancestors called their home could have been a matter of life or death for them? Today, everyone knows that rich people generally live longer than poor people because they can afford money to leverage hi-tech medical  facilities, but what about people 2 Centuries ago?

We will try to find answers to these questions and discuss if our main claim holds good by exploring this amazing interactive visualization.The visualization which we are about to discuss reviews data from 200 countries and compares life expectancy vs wealth for the past 200 years. The vertical axis shows the average life span in each country ranging from 25 – 85 years, where high up= long lives= good health, to the bottom= shorter life=sick. The horizontal axis shows the average income per person (GDP per capita) expressed in dollars per person per year, where the right=rich and to the left=poor. It’s interesting to see the usage of bubble chart for this, which is primarily used when you represent data that has three or more data series (In this case income, life span and size of the population) and each containing a set of values.

UNDERSTANDING THE DATA:

Let’s explore the visualization and understand it better. On the first look we can figure out that each country in the world is a bubble,the size of the bubbles represent the population size,color represents regions of the world (see on top right side). We start with circa 1800, all the countries had life expectancy less than 45 years and an income less than $4500. We can see that the United Kingdoms & Netherlands were among the richest countries but people in there had short lives. Underdeveloped healthcare systems and poor sanitation attributes to some of the reasons why all the countries had shorter life span and most importantly these acts as a warrant to our claim.

Now as we click ‘play’ the years start to roll in the world. Slowly income start to increase mainly in Europe and North America because of industrial revolution. As a result, they pulled away from the rest of the world. BUT, surprisingly health didn’t get much better. In 1900, only western countries were getting richer and richer and became healthier and healthier. Between World War I (1914) and WWII (1945) the difference between the rich and poor countries increase and it’s only after the WWII that most countries started to change in terms of wealth. The Arab countries became the richest and countries like China and India prosper as a result of their emerging economic growth.

Now in 2017, we observe a continuous world with high income countries (Qatar,Norway,USA) having a high life span and low income countries (Ethiopia, Niger, Liberia) have a lower life span, but interestingly all the countries are estimated to have more than 45 years of life expectancy, which only happens to be the maximum life of people in 1800. Though the difference between high income countries and low income countries are huge but their respective citizen’s longevity have come up significantly.

DRAWBACKS:

Undoubtedly the visualization is amazing in itself, but there are few snags which can alter the statistics if taken into consideration. First, the data collected are only with respect to inter-countries. But what it doesn’t include is the scope to look at the differences of incomes within the regions of a given country which would give insights to it’s growth/downfall. Second, while talking about population size of any country we only take into account it’s current citizens but there is a significant inflow of immigrants in these country every year contributing towards the economy. So there is a high probability that it might give us a different picture altogether.

FROM A CRITIQUE’S VIEWPOINT: 

The number of different parameters presented on the interactive dashboard are overwhelming. For a new user it becomes hard and confusing, instead a simple drop down could be introduced to give the audience the flexibility to play around with their desired set of parameters. The vertical lines on the chart needs to be even spaced and the text for year should be at the top to avoid any kind of visual conflict. Also, while toggling on the ‘Map’ tab, it gives us an elliptical view of the globe and the bubble of each country doesn’t sync very well with their respective geographical location. This can be eliminated by displaying a flat world map view and being accurate about the geographical locations.

ALTERNATIVE APPROACH/MODIFICATION:

Though it’s visually appealing there are certain hiccups with this bubble charts visualization as well. It can be further enhanced and made simpler by adopting certain techniques.

1. The bubbles are opaque in nature creating a problem to clearly figure out countries with smaller population size. So, as an alternative I would recommend to use translucency and highlighting the boundaries of the bubble. These are powerful tools for dealing with over plotting, as you can see this in below visualization.
2. As we discussed earlier that the visualization doesn’t show the differences within a country, It can be modified by introducing an additional feature in which by selecting a country say United States, it will give you an overview of all the data values for the 50 states along with an appropriate color contrast. The modified version looks like the below visualization. 

CONCLUSION:

I feel that this kind of visualization is really helpful  when conveying a large amount of numeric information quickly to your audience but at the same time ensuring that viewers are visually literate. An important part of bubble chart visualization is to make sure that it is clear what each element of the chart means – color, circumference, how it fits on the scale otherwise the whole meaning can be lost. Similar approach/viz can be an advantage for organizations to analyze their financial sales with respect to their customer base. It will help them to come up with business metrics and promotional plans for their consumers.

Reference: Harvard Gazette, The New York Times 1, The New York Times 2, MIT News

Bikram Patnaik

Google Play Music timeline 

‘What would you do if you want to listen to an old classic country music on Friday evening after a hectic day at work?‘

Generally we turn up to various websites, blogs, TV channels and radio which lists the top 20 music tracks/albums of a particular year but, Google Play Music’s makes it easy with it’s new Music Timeline visualization which gives us a bird eye view of our past musical favorites and gives us a chance to revisit them. It helps us visualize which music has stood the test of time, and how genres and artists have risen and fallen in popularity.

The Music Timeline uses data from Google Play Music users’ libraries to categorize artists by genre, and the genres are then subdivided. What it provides, then, is a rough-and-ready map of the popularity of genres and artists over the years.The X-axis shows us the transition of time from mid-50’s to 2010 while the Y-axis scales the popularity of that particular genre. Here, the visualization uses stacked area chart which are usually used in situations when we need to display some changes in time, when it’s important to show that those values in a sum form a whole. For all the music loving audience there can’t be a better way of representation than this.

UNDERSTANDING THE DATA:

Let’s dive deeper by understanding it’s working dynamics. When you glance across the timeline it gives a soothing treat to the eyes with it’s subtle color combinations. You can figure out straight away that during the early 50’s ‘Jazz’ & ‘Vocal/Easy-Listening’ genre were very popular among people. But as time elapsed ‘Rock’ and ‘Pop’ culture picked up the pace and overshadowed other genre. In the early 80’s ‘Jazz’ along with other genre reached their threshold and people developed a different taste of music. Emergence of new artists like Snoop dog,Eminem and Nirvana resulted in the up rise of ‘Alternative/Indie’ and ‘Hip-hop/Rap’ culture and made it a craze.

Once you’ve drilled down to your selected genre, the timeline takes the form distinctive audio wave showing the flow of popular individual artists/bands and displays a short bio and relevant albums. For example, by clicking on the Pop stripe, we can see the combination from ’50s Pop to ’60s Pop to Adult contemporary within the growth of the overall genre, as well as some of the most popular artists that composed each sub genre. This helps audiophiles to choose the artist of their choice and buy relevant songs.

Another interesting feature is that we can search for a particular artist to see the trajectory of their career through the decades. Let’s say, Michael Jackson who started his music career in 70’s but didn’t hit his stardom until the release of his famous album Thriller and Bad in 80’s after which his legacy still continues till date. The same feature also applies to music albums.

DRAWBACKS:

Now talking about the loop holes in this visualization, the data collected is only restricted to Google Play Music user’s libraries and doesn’t take into account users from other music friendly platforms like Apple music, Spotify,Pandora or Sound-cloud. If we collate the data from all sources, there is a high probability that it might give us a different picture altogether. More ever, the very existence of Google Play Music in 1950 or it’s use by the old generation people, who mostly prefer traditional ways (cassette players and cd-player) of listening to music is questionable. I leave that up to you to decide. But certainly the variety and easy usability of Music Timeline over powers it’s flaws.

FROM A CRITIQUE’S VIEWPOINT:

While exploring a particular genre say ‘ROCK’ and diving deep into it’s sub-genres, if you examine closely the word ‘ROCK’ is embossed in the background throughout the entire audio wave format. This seems to be overwhelming specially when you have multiple sub-divisions each highlighting it’s own name. A color contrast is much needed to distinctly identify the sub-categories. Lastly, the font size needs to be standardized across the timeline to reduce the probabilities of missing out on words/texts.

ALTERNATIVE APPROACH:

Though it’s visually appealing there are certain hiccups with stacked area charts as well. Ideally one should be able to interpret each individual series by its height, but unfortunately most interpret the curve of the top of the area as indicating quantity ( In line graphs). So, as an alternative I would recommend individual line charts, with an additional line in a stark color for the total or we can use interactivity and gray out series in the background, such as in this amazing visualization of housing prices from THE New York Times 

CONCLUSION:

I feel that this kind of visualization is really helpful  when we have to organize huge data sets (live data) across a defined time-line. Similar approach/viz can be an advantage for organizations to analyze their product sales along with their respective popularity. It will allow them to come up with business metrics targeting valuable customers.

Reference:

https://www.theguardian.com/music/musicblog/2014/jan/17/google-play-music-timeline-punksoul

The NYT