UPCOMING WEBINAR: Cluster analysis: Classic, powerful, and still worth knowing Counts Analysis is a method used for analyzing MaxDiff data which counts the number each item is selected as Best and subtracts the number each item is selected as Worst.
This article takes you through how to perform this analysis technique in Displayr.
Requirements
- A Displayr document with a data set that contains a set of MaxDiff variables which store the Best and Worst selections.
- In this example, we have collected data from a single version design with 6 questions and 10 technology brands as alternatives. The variables have been set up as one Best variable per question and one Worst variable per question. Below is a raw data view of these variables.
Method
1. To begin with, your MaxDiff Best and Worst variables should not be grouped together. If they are, you will need to first select them in the Data Sources tree and right-click > Split.
2. If your MaxDiff variables are already split as separate variables, select them in the Data Sources tree, otherwise you can skip this step.
3. Click > Filter > Filters From Selected Data.
4. While the new variables are highlighted, untick Usable as a filter in the object inspector.
5. Next, select only the Best filter variable sets and click > Calculate Across Variables > Sum.
6. Right-click > Rename these variables to change the variable set Label to Best.
7. Now select only the Worst filter variable sets and click > Calculate Across Variables > Sum.
8. Right-click > Rename these variables to change the variable set Label to Worst.
9. Select both variable sets and click > Convert To > Grid.
10. Drag your new Best + Worst variable set onto your Page as a table.
11. Change Data > Statistics to Sum.
12. In the object inspector, click Data > Rules > + > Table Computations > Difference Between Pair of Columns.
13. Select Best in the Column 1 field and Worst in the Column 2 field.
14. [OPTIONAL] Untick Label new column automatically and enter the preferred label in New column label.
15. [OPTIONAL] Tick Replace table with difference if you wish to hide the other columns.