Depending on the nature of your project, a simple A vs B comparison may not be sufficiently robust to produce the insights you are looking for. When dealing with a group of three or more data sets, NOT comparisons are a useful tool for analysing the data.
In this article:
What is a NOT comparison?
'NOT comparisons' are commonly used when a project involves more than two data sets as a way of identifying what makes a particular data set unique in comparison to all of the others grouped together.
NOT comparisons involve comparing a specific data set against a group of others combined and is constructed as (X) vs (NOT X).
Consider a scenario where you have four data sets to analyse:
[A] vs [B+C+D]
[B] vs [A+C+D]
[C] vs [A+B+D]
[D] vs [A+B+C]
Common use cases
This approach can be useful whenever you are working with more than two data sets. Some common examples include:
Analysing one brand against a group of competitors
Isolating a particular audience group (e.g. age group, geographic group) and comparing it against all others
Comparing NPS responses from promoters against detractors and neutrals
Creating NOT comparisons
Navigate to the Data Library and open the relevant project folder (or create a new project and upload data).
To create a NOT comparison including only a selection of the data sets in your project, click 'New folder' then select the relevant data sets and click the move icon from the bulk actions menu at the top of the screen.
Hover over the single data set you want to isolate from the others and click the two circles icon to compare against all others.
You will be brought to a preview of the comparison. Click 'Save' to save this comparison to the project.
The combined data set will be saved within your Data Library using the format 'NOT xxxxx', and can be used in other comparisons as needed.
To prevent the analysis of duplicated data when creating additional NOT comparisons, any NOT data sets created previously will be excluded.