Weights are used to adjust the proportionality of the data records. There are several ways that weights can be used for analysis. You can have a default weight set in the project, auto weights, and you can manually select a weight.
In this article
1. Default Weight
Each project can have one weight set as the default weight Learn more about weights.. This means that whenever an analysis is created, the weight set as the default is applied.
If a default weight is set, all analyses will be weighted. A default weight item is indicated with a blue dot. Other weight items may also be available in the project.
-
Default Weight
-
Weight
When creating an analysis when there is a default weight in the project, the weight label displays in the interactive title.
- Create an analysis
- The default weight label displays in the interactive title
If there is no default weight in the project, your analysis remains unweighted and you can manually apply a weight using the weight drop zone.
2. Weight Drop Zone
The Weight drop zone allows you to apply a weight to your analysis. The weight drop zone only accepts items that are defined as a weight.
Apply a New Weight to the Analysis
If there are no weights already applied to an analysis, or you want to override the applied weight, you can drag and drop a weight variable into the weight drop zone of the analysis.
After applying the new weight, you can use the interactive title to switch between the newly applied weight and the default weight.
- Create your analysis table
- From the project tree, select and drag an alternative weight across to the Weight drop zone
- The interactive title displays the weight applied in the analysis
- You can switch back to the default weight
Once you change a default weight by dragging in a different weight, you can click on the weight item in the interactive title to remove it and view the unweighted analysis.
- After replacing the default weight, click the weight item in the interactive title
- Click the selected weight to remove it
3. Auto Weights
Specific variables may have their own auto weight identified with the blue i icon. This means that whenever that variable is included in an analysis, the auto weight is also applied to the analysis.
It is possible to include more than one variable with different auto weights together in either the across or down of the analysis. When you drag two axes into the same drop zone, each with their own auto weight, each retains their weight and you'll see the label Multiple Weights in the interactive title.
In the following auto weight examples, two auto weights have been applied:
- Gender - Booster Weight is the auto weight
- Travel Style (Grouped) - Weight is the auto weight
Here, we can see the Multiple Weights label when two axes with auto weights are in the down drop zone.
- The blue i icon next to the variable in the project tree shows the auto weight
- If you drag multiple items with auto weights to the same dimension, Multiple Weights shows in the interactive title
Auto weights are not applied in the following circumstances:
Auto weights do not apply in both the down and across dimensions
You cannot have different weights applied to the across and down drop zones. If items with auto weights are selected in different dimensions of the analysis, the weight(s) associated with the down variable(s) is applied. Any auto weights for variables in the across dimension are not applied.
In the following GIF, Gender is in the down drop zone and Travel Style (Grouped) is in the across drop zone. Only the auto weight for Gender applies.
- When Travel Style is dragged into the across drop zone, the auto weight displays.
- When Gender is dragged into the down drop zone, the Gender auto weight takes precedence over the across drop zone.
In nested analyses, Auto weights are not applied to the outer nest
In a nested analysis, auto weights in the outer nest variables are not currently supported. Only variables in the innermost level of a nested analysis will retain the auto weight.
In the example of a nested analysis, when Travel Style (Grouped) is nested as the inner nest in the analysis, Travel Style (Grouped)'s auto weight, Weight, is applied. But when Travel Style (Grouped) is nested as the outer nest in the analysis, the auto weight, for Gender, Booster Weight, takes precedence.
- With Travel Style as the innermost nest, the auto weight, Weight, applies
- With Gender as the innermost nest, the auto weight, Booster Weight, applies
When you drag a weight to the measure drop zone, the weight overrides any auto weights included in the analysis
4. Weights in a Multi-Level Project
In a multi-level project, each level may have a separate default weight applied. An analysis table includes the correct weight from each level.
If you drag and drop a weight variable in a multi-level project into the weight drop zone, the new weight will apply to all levels, overriding all the default weights.
Learn more about multi-level weights.
Where to from here?
Learn more about analysis: