Adding Global Per-Record Weights
Applying a Single Weighting Scheme
Per-record weighting requires that each respondent has a float-type question that contains the record weight to be applied. If you are using Crosstabs, you can create a weighting scheme to use via the RIM Weighting Tool. The weighting scheme created this way will be stored in a virtual question called
The above question contains the per-respondent weighting coefficients, and can be applied as a weight to a specific segment in your dashboard. In the below example, the
nweight question is used to apply a 50:50 gender split for your data.
weight nweight.val table id=db-1 Gender Weights rows s1.r1-r2
The above example produces the following result:
Working with Multiple Weighting Schemes
While a single weighting scheme can be easily applied by referencing
nweight.val, multiple weighting schemes created via Crosstabs change the way your
nweight question works. Instead of being a number question with no explicit rows, it will have one row created for each weighting scheme you create, with each row taking the weighting scheme name as its label:
With that in mind, you cannot apply multiple global weights to a dashboard. If you specify multiple
weight attributes, this will result in only the last defined weight to be applied.
For example, if you would like an equal distribution across both age and gender, you would use the following:
weight nweight.AgeGroups.val weight nweight.Gender.val table id=db-1 Gender Weights rows s1.r1-r2 table id=db-2 Age Weights rows GEN.r1-r3
This will produce the following result, where only your gender weight is applied, as it was defined last:
Adding Per-Segment Weights
weight attribute can also be applied to a specific segment in your banner split to apply per-record weighting in a specific table.
Note: Click here to learn more about Creating Dashboard Banners.
weight keyword is only available in compat level "2 (faster)" dashboards.
table id=db-1 Multiple Weighting schemes - Gender banner.local segment weight=nweight.AgeGroups.val qualified Age Weighting segment weight=nweight.Gender.val qualified Gender Weighting rows s1.r1-r2 table id=db-2 Multiple Weighting schemes - Age banner.local segment weight=nweight.AgeGroups.val qualified Age Weighting segment weight=nweight.Gender.val qualified Gender Weighting rows GEN.r1-r3
The above example allows you to apply different weighting schemes to the different segments that you’ve defined. As displayed in the table below, only the "Gender Weight" segment is affected in the first table, as you are applying a weighting scheme based on your gender question, while the "Age Weighting" segment in the second table has an equal distribution:
Note: Configuring weights on a segment will override any global weight configuration. This is mostly used when you want to compare weighted and unweighted data for the same survey.
Additionally, you can combine weighted and unweighted segments together to compare weighted and unweighted totals:
table id=db-1 Weighted vs. Unweighted Counts banner.local segment qualified Unweighted Data segment weight=nweight.Gender.val qualified Weighted Data total rows s1.r1-r2
This produces the following table: