Suicide rate analysis from 1985 to 2016

I recently did an analysis to interpret the pattern of suicides committed across all age groups in certain countries from the year 1985 to 2016. Even though suicides are one of the leading causes of deaths, it is still very under reported in many countries which makes it impossible to do an accurate analysis across the world. The data I have comprises of 30 countries across different age groups from 15–75 + years.

Initially, I tried to find the correlation between the gdp (per capita) of the country and the number of suicides. As the population of each countries varies, I decided it would be better to take suicides per 100k population instead of the total number of suicides.

The reason behind plotting the correlation between gdp and the number of suicides being plotted for these years (1985, 1991, 1997, 2004, 2009, 2015) is that the pattern is almost similar during the other years and leaving a gap of 6 years makes it easier to visualize the progress.

It is clearly evident that as the gdp increases, the number of suicides committed decreases. The country with the maximum number of suicides somehow always falls in the low gdp category.

In order to find the pattern of suicide rates across the world, I decided to do a heat map.

This is the list of top 20 countries based on suicide rates. I could do a heat map only from 1995 because there were a lot of missing values in the early 80s and 90s mainly because the number of suicides were very under reported across the world. Overall, the suicides have been clearly decreasing in most of the countries. Hungary, Lithuania and Russian Federation have had the maximum number of suicides but even those countries have seen a decline in suicide rates but there are exceptional countries like Suriname, Republic of Korea and Guyana where the suicide rates have been increasing. Even though Republic of Korea has had as little as 65 number of suicides in 1995, after 2000 the suicide rates have been increased to 600s.

In addition to this, I was interested in finding the pattern of the number of suicides committed by male and female over the years. By taking the help of time series plot which will represent suicide rate pattern over the years, this is what I arrived at.

There is a significant gap in the number of suicides committed by male and female over the years across the world. The number of suicides by males went on rising from 1985 to 1995 but the female suicides have been almost linear throughout. In the late 2000s, there has been a drastic decrease in the number of suicides committed by males which is almost as low the female suicides.

Then I did a hypothesis testing to test the mean number of suicides between 15–24 year olds and 25–34 year olds from the year 2015 to get a recent analysis on the same. Here the null hypothesis is that the suicide rates between these age groups is the same. After performing the Welch two sample t-test and plotting the t-statistics plot which is shown below

Every time I run the test with random samples from the data, the T calc always falls in the acceptance region and hence, we fail to reject the null hypothesis implying that there is no significant difference in the suicide rates between 15–24 and 25–34 year olds since 2015.

It was really surprising when I had a look at the data set and happen to note that the number of suicides committed by the population aged above 75 were very significant in number. Hence, I decided to calculate the confidence interval for the difference in proportion of suicides committed by male and female aged 75 and above with 95 % confidence. First, I considered two populations of the number of suicides committed by male and by female in the year 2000 to 2016. I performed the t test to find the lower and upper limits for the difference between two independent proportions, say the proportion of 75+ males who committed suicide among the whole population of number of male suicides and the proportion of 75+ females who committed suicide among the whole population of number of female suicides.

For every random sample, the lower limit and the upper limit of the confidence interval had alternating signs implying that there is no significant difference in the proportion of suicides committed by male and female aged 75 and above.

Even though the suicides have been decreasing in most of the countries, suicide deaths are still one of the leading causes of mortality across the world and also the most under reported. We cannot deny that the awareness of mental health have been increasing which seems to be an impact in the suicide rates. The more we talk about mental health, the more this world become a better place.