Look at the landscape of your data
With correspondence analysis it is easy to find the underlying structure of your data. Consider a table like this one:
Here people were asked what attributes they would associate with which brand. Given the number of brands and attributes it is very hard to draw conclusions from this table.
This is where correspondence analysis comes in. It takes the information from the table and creates a map like this one:
Now we can get a better understanding of the data with some glances at the map. It is important to understand that positioning plays an important role with correspondence analysis. We can rightly assume that Google has a high association with “Fun”. But it would be wrong to say that Samsung associates highly with “Worth what you pay for”.
To explain this consider the same map where I drew two arrows:
The arrows help us understand how correspondence analysis works. If we draw arrows from the origin (the center) of the map to the two points that interest us, two things are important: the length of the arrows and the angle the two arrows form. The longer the arrows and the sharper the angle they form, the greater the association between these two points.
This explains why Google and “Fun” have a higher association (63%) than Samsung and “Worth what you pay for” (43%) although they are closer to each other than Google and “Fun”.
Get a map of your data to quickly find the insights you are looking for
Correspondence analyses have a wide range of use cases. In general, if you have a table that is to hard to read, a map produced by a correspondence analysis can be produced and be helpful. There are many ways to visualize a correspondence analysis. What I showed you here is just the beginning.
If you decide to use correspondence analysis in your project some things – e.g. how you ask questions specifically – are important to consider. I’m happy to help you with this, not just with the computation and presentation itself.
I can send you a customized quote for your project at anytime. Please just email me your specifics and we can set up a kickoff call or exchange further details via mail: email@example.com