Using Table Dimension Transforms

1: Summary

The table dimension transform creates a table where each row represents a dimension from a specified dimensionGroup. It’s commonly used to calculate and output scores (like averages) or categorical values for each dimension.

This is useful for dashboards, scoring models, or any structured reporting based on grouped survey content.

2: Declaring the Table

table dimension #yourTableName {
  parent: .yourParentTable:
  dimensionGroup: @yourDimensionGroup
  action create {
    // variables here
  }
}

2.1: Key Properties

Property	Description
parent	The table where the new dimension table will be created (typically .response: or a loop level).
dimensionGroup	A reference to a dimensionGroup entity. Determines which dimensions will be included.

3: Creating Variables

Inside action create, you define one or more variables to be calculated per dimension:

action create {
  variable numeric #dimensionScoreValue {
    value: dimensionAvg()
    required: true
    totalDigits: 7
    decimals: 6
  }
}

3.1: Example: Three-Point Distribution

You can also recode the average score into a simpler scale:

variable singleChoice #threePoint {
  value: recode(dimensionScore:dimensionScoreValue, @dimension3point)
  label: "Distribution"
  required: true
  option code { code: "3", label: "Unfavorable" }
  option code { code: "2", label: "Neutral" }
  option code { code: "1", label: "Favorable" }
}

4: Using numericlist Fields

You can also define numericlist variables to break down count distributions per dimension:

variable numericlist #favCounts {
  totalDigits: 5
  decimals: 0
 
  field {
    code: "3"
    label: "Unfavorable"
    value: sum(iif(score(dimensionItems()) <= 2, 1, 0))
  }
  // Additional fields...
}

5: Example: Full Dimension Table

table dimension #dg1_dimensionScore {
  parent: .response:
  dimensionGroup: @dg1
  action create {
    variable numeric #dimensionScoreValue {
      value: dimensionAvg()
      required: true
      totalDigits: 7
      decimals: 6
    }
      variable singleChoice #threePoint {
      value: recode(dimensionScore:dimensionScoreValue, @dimension3point)
      label: "Distribution"
      required: true
      option code { code: "3", label: "Unfavorable" }
      option code { code: "2", label: "Neutral" }
      option code { code: "1", label: "Favorable" }
    }
       variable numericlist #favCounts {
      totalDigits: 5
      decimals: 0
      field { code: "3", label: "Unfavorable", value: sum(iif(score(dimensionItems()) <= 2, 1, 0)) }
      field { code: "2", label: "Neutral", value: sum(iif(score(dimensionItems()) = 3, 1, 0)) }
      field { code: "1", label: "Favorable", value: sum(iif(score(dimensionItems()) >= 4, 1, 0)) }
    }
  }
}

6: Best Practices

Always declare dimensionAvg()first in the variable list. This ensures compound dimensions are calculated correctly.

Keep dimensionGroup logic centralized to ensure consistency across tables.

Use recode() and numericlist for simplified reporting.