Re-mapping Female Homicides; How maps can change what we see

There were exactly 1675 female homicides between 2010 and 2017. Meanwhile, maps, as a part of our struggle to understand these never ending murders, are always a part of our research. Maps have the power to demonstrate the prevalence of these murders and connection between them and their locations.

I will try to demonstrate why the most common technique that is used to map female homicides can misrepresent the magnitude of the issue and I will propose an alternative way to map these murders to overcome this misrepresentation.

Cluster Map with Total Numbers of Murders

Cluster Map with Total Numbers of Murders

We see the most common way to map female homicides (in this case in Turkey) in this cluster map on the left. The map shows the number of murdered women in each city between 2010 and 2017. Below we see the representation of the same map in a polygon style. The polygon map shows us the same information with color codings. The total number of murders are divided into 4 equal categories. 

Polygon Map with Total Numbers of Murders

> 172 to 230
> 115 to 172
> 58 to 115
> 1 to 58

When using this four intervals, we see that 6 cities are distinguished by their total number of murders. These cities have > 58 total murder number. The rest of the cities fall under < 58 category.

Until here we only saw two different map styles (cluster and polygon) that uses the most common way of mapping murdered women in Turkey. Each city is represented by the total number of the murders. Does this seem fair enough to you? Well for most of us it does. It delivers the necessary information to the public. But what if a different mapping approach can change how we perceive female homicides? 

In order to really understand the magnitude of female homicides I did a simple calculation to analyse murders within their place of occurance. So I divided the total number of murders in a city to the city population. This calculation shows us that 39 cities are distinguished from the rest. Compared to 6 cities from the previous tecnique this is a huge difference.

Results for "total numbers of murders"

Results for "total numbers of murders"

Results for "murder / city population"

Results for "murder / city population"

To make things clear lets look at the results in two mapping approaches and compare them. In the first mapping technique we see that the following cities has the highest number of murders. As the number of murders are higher in Istanbul (230) it is mapped with darkest color in our first polygon map (> 172 to 230).

But when we analyse each city within themselves we see that Iğdır has the highest number of murders considering the population sizes. Despite having 14 murders in total, this number is more significant than we think. With this technique, our polygon map (that has 4 categories) will have 39 cities collered that will fall under the top three category. Moreover, we observe that Istanbul is no longer at the top of the list but it's the 54'th city in the murder/population ratio.

Click on the image below to check the interactive map

Polygon Map of Murder Ratio to City Population

Polygon Map of Murder Ratio to City Population

Why is this relevant?

You may ask why we should care about how we map the social issues. The fundamental reasons for visualisation and map making, aside from opening up the data to the public and making it understandable, are to stress the importance and the magnitude of a specific event and, to show us which locations need our attention on the way to demand better social policies.

As we always stress, murdered women are not merely numbers for us. Each death is equally painful for us. However, we need better and more accurate mapping techniques in order to understand the magnitude of any social situation.

Data

All data is retreived from kadıncinayetleri.org; A powerful project conducted by beautiful people to show us the magnitude of the female homicides.