Comparative Crosstabs
Engine provides the ability to compare the results of one crosstab with another - perhaps to analyse how well represented a group of records is in one population compared to another. This kind of crosstab is referred to as a comparative crosstab, and relies on the fact that crosstabs can be calculated for subsets of data.
Comparative crosstabs can be created for both Standard and Multi-dimensional crosstabs.
Applying a filter
Crosstab results can be easily calculated for different sub-sets of data through the application of a filter. By applying more than one filter to a crosstab it becomes possible to create a comparative data set.
For example, a filter of "Gender='Male'" can be applied to a crosstab of Occupation by AgeBand. You can see in the figure below that the number of directors who are 65 and over has decreased from 178 for both genders, to 137. The number 137 represents the number of Male Directors who are 65 and over. The crosstab is now displaying data for the TARGET group - the group with the filter applied.
| RESOLUTION TABLE | [DEMO].[CUSTOMER] |
| Y-AXIS | [DEMO].[CUSTOMER].[OCCUPATION] |
| X-AXIS | [DEMO].[CUSTOMER].[AGEBAND] |
| TARGET FILTER | GENDER='M' |
The display mode of the crosstab can now be changed from the default COUNT to INDEX.
Index values greater than 1 indicate that the group of records contained in the intersection are over represented in the target group when compared to the base group, where the base group is the unfiltered group. Index values less than 1 indicate that the group is under represented.
In this example, directors over 65 have an index of 1.26. This is greater than 1 and therefore indicates that there is a higher percentage of directors over 65 in the target group than the base group. In other words, directors over 65 are more likely to be male:
