Despite the obvious advantages of this approach such as high result precision and compact layout, it has it’s drawbacks.
Those who analyse large amounts of data in crosstabs every day know how tiresome and unrewarding this job is. In order to draw a conclusion whether or not there is a relationship in the data, they have to read, process, and compare many numbers, which is very time-consuming and requires absolute focus. Can it be made easier without compromising the accuracy of the crosstab?
The contingency map allows you to keep the table layout and replace or supplement the number in each cell with a graphical representation. By further selecting colour and colour intensity to represent numerical values, cells may be filled with colours the intensity of which may reflect the statistic in the cell. The direction of colour intensity may also be determined such that cell colours will depend on the selected direction of percentage ordering in the table.
The concept behind the contingency map is taken from the popular ‘heat map’ visualisation and is the perfect compliment to the crosstab rendering them much easier to interpret.
The contingency map above presents customer satisfaction data for a lunch bar. The selected statistic was ‘percentage in column’. With a glimpse at the intensity of the green in the cells you can quickly tell that there is a strong correlation between the waiting time and the readiness to recommend the place to friends. The shorter the customer has to wait, the more readily they will recommend the lunch bar.
Additionally, the visualisation outside the box shows the frequency distribution of both analysed variables in the form of bars. This way we know which responses were selected most often by the participants.