Infotools Harmoni has a lot of flexibility to define the base for a result. The default is to use the Total. Learn more about Total Bases.However, if the Total is not what you want, you can apply a different base either by using apply base Learn more about how to apply a base in the project tree. in the project design or by setting the percentage base in the analysis. Learn more about analysis drop zones.
Percentage base drop zones allow an item or variable item to be specifically applied as the base for the analysis.
This option is useful when:
- You want to change the default down percentages to across percentages or vice versa or
- You want the analysis to be focused on a specific group and
- You want the base to be the same regardless of the direction of the table.
In this article
- Percentage base on down or across
- Percentage base and filters
- Percentage base on Cell
- Percentage base on Total
1. Percentage base on down or across
In an analysis with at least one variable down and one across, the default percentage base for calculations is the down dimension. When viewing results, there are times where you might prefer to base these percentages on the across.
You can interact directly with the label that indicates the current direction (what the percentages are based on) and it allows you to select another from the available list.
Click on percentages based on, and select the relevant metric from the drop-down menu (i.e. your down dimension in a cross tab).
- Down Percentages calculate the percentage that each cell represents as a proportion of the total for each column in the table. The base for down percentages is the value in the Total row at the top of the table.
- Across Percentages calculate the percentages that each item represents as a proportion of the total for each row in the table. The base for across percentages is the value in the Total column to the left of the table.
2. Percentage base and filters
When using a filter, Learn more about analysis with a filter. the filter selected applies to the entire analysis, that is the Total counts (numerator) and the base (denominator). However, when using a filter on an analysis that has a percentage base applied the following applies:
-
Filters default to applying only to the Total counts (numerator) but not the base (denominator). The base is determined by your selected axis or element.
- Apply filter to base option - filters apply to the entire analysis, that is, the Total counts (numerator) and the base (denominator). Note that the filter has to be a different item to that in the percentage base.
3. Percentage base on Cell
Percentage base on cell is useful when you want to apply a non-plotted variable (i.e. not one that is already in the across or down) to the base of each cell in the analysis.
The Apply Filter to Base feature provides the ability to choose whether the percentage base selection should exclude records that satisfy the filter selection.
Percentage base on cell option only becomes active when you already have an analysis with items in the down and across.
When you use the percentage base on cell analysis drop zone, each cell in the table has a different base. You can drag either an axis or an element into the cell analysis drop zone.
-
If an axis is dragged in, the cell base is the count of records in the cell AND the selected percentage base axis. Selecting an axis doesn't automatically add a filter.
-
If an element is dragged in, the cell base is the count of records in the cell AND the selected element. Selecting an element will add the selected element as a filter in the analysis.
-
When dragging an axis or element into the percentage base cell, percentages will likely be 100%. This is because both the counts and base reflect the intersection of all selected items (i.e. the count of records in the cell AND the selected percentage base axis or the selected element).
-
Selecting the same axis or element as a filter will filter the Total counts (numerator) but not the base (denominator). The base is the count of records in the cell AND the selected percentage base axis or element.
-
Any additional filters applied to the analysis are not applied to the base unless you use the apply filter to base option.
Example
Q. Amongst all traveler types, what proportion of females who have given a Top 2 answer to Overall Satisfaction are independent travelers?
Given that the focus is a specific group within the sample i.e. gave a response to the Traveler Type variable, we can use the cell analysis drop zone for this analysis, adding a filter into the analysis to specifically target Independent Travelers, while keeping the base of all traveler types.
Note, if a filter element exists in the same variable as used in the percentage base analysis drop zone, the percentage base is not filtered by the filter item, as this would always result in 100%. Instead, the filter is only applied to the cell count.
A. Amongst all traveler types, 63% of females who gave a Top 2 answer to Overall Satisfaction (n=365) are also Independent Travelers (n=230).
Q. Amongst all traveler types, what proportion of females from Australia who have given a Top 2 answer to Overall Satisfaction are independent travelers?
For this example, we will add into the analysis another filter, in this case, the element Australia. Notice we have the option to apply a filter to the base as well.
- Apply filter to Base off:
Amongst those with a traveler type response, 18% of females who gave a Top 2 answer to Overall Satisfaction (n=365) are also Independent Travelers from Australia (n=65).
- Apply filter to Base on:
Amongst those with a traveler type response, 62% of females from Australia who gave a Top 2 answer to Overall Satisfaction (n=105) are also Independent Travelers from Australia (n=65).
A. Amongst traveler types from Australia, 62% of females who gave a Top 2 answer to Overall Satisfaction (n=105), are also Independent Travelers(n=65).
Scenario |
Total (qualified) |
Counts |
Base |
Percentages Calculation |
What does this mean? |
Percentage Base Cell (Traveler Type - Axis) |
Total (qualified) =474* The total is Females who gave a response to the Overall Satisfaction Question. | 365* The counts are the intersection of Traveler Type, Satisfaction and Gender. (Females who gave a Top 2 response and are included in Traveler Type) | 365* The base is different from the total as the percentage base takes priority(Females who gave a Top 2 response and are included in Traveler Type) | Traveler Type percentages based on overall Satisfaction and Gender and Traveler Type |
100%. Both the counts and base reflect the intersection of all selected items (Females who gave a Top 2 response and have a Traveler Type). |
Percentage Base Cell + Filter (Traveler Type - Axis) filter: Traveler type |
Total (qualified) =306* The total is filtered by the selected Axis/Element i.e Traveler Type: Independent |
230* The counts are the intersection of Traveler Type, Satisfaction and Gender filtered by Independent Traveler. |
365* The base is different from the total as the percentage base takes priority (i.e. Traveler Type). The filter has no effect on the base. |
Traveler Type percentages based on overall Satisfaction and Gender and Traveler Type filtered by: Traveler type: Independent |
Amongst traveler types 63% of females who gave a Top 2 answer to Overall Satisfaction and have Traveler Type are Independent Travelers. Top 2= |
Percentage Base Cell + More than one Filter (Traveler Type - Axis) filter: Traveler type and Country: Australia No Filter to Base |
Total (qualified)=91* The total is filtered by the selected Axis/Element i.e Traveler Type: Independent and Country: Australia. |
65* The counts are the intersection of Traveler Type, Satisfaction and Gender filtered by Independent Traveler and Australia. |
365* The base is different from the total as the percentage base takes priority (i.e. Traveler Type). The filter has no effect on the base. |
Traveler Type percentages based on overall Satisfaction and Gender and Traveler Type filtered by: Traveler type: Independent and Country: Australia base filter: Off |
Amongst traveler types 18% of females who gave a Top 2 answer to Overall Satisfaction and have a Traveler type are Independent Travelers from Australia. Top 2= |
Percentage Base Cell + More than one Filters (Traveler Type - Axis) filter: Traveler type: Independent and Country: Australia Apply Filter to Base) |
Total (qualified)=91* The total is filtered by the selected Axis/Element i.e Traveler Type: Independent and Country: Australia |
65* The counts are the intersection of Traveler Type, Satisfaction and Gender filtered by Independent Traveler and Australia. |
105* The base is different from the total as the percentage base takes priority (i.e. Traveler Type). The base has also been filtered. |
Traveler Type percentages based on overall Satisfaction and Gender and Traveler Type filtered by: Traveler type: Independent and Country: Australia base filtered by: Country: Australia |
Amongst traveler types 62% of females from Australia who gave a Top 2 answer to Overall Satisfaction and have a Traveler type are Independent Travelers. Top 2=
|
4. Percentage base on Total
Percentage base on Total allows you to create an analysis where all cells use a base that is neither the down nor across variable.
Percentage base on total option only becomes active when you already have an analysis with items in the down or across.
When you drag an axis or element into the percentage base on total analysis drop zone, all the cells in the table get the same base.
-
If an axis is dragged in, the table base is all records with a response in the dragged axis. Selecting an axis doesn't automatically add a filter.
-
If an element is dragged in, the table base is all records with a response to that element. Selecting an element adds the selected element as a filter in the analysis.
-
Any additional filters applied to the analysis are not applied to the base unless you use the apply filter to base option.
Example
Q. Amongst all independent travelers who gave a response to Overall Satisfaction and Gender, what proportion are females who have given a Top 2 answer to Overall Satisfaction?
Given that the focus is a specific group within the sample i.e. Independent Travelers, we can use the analysis drop zones for this analysis. In this case, we will use the element Independent.
A. Amongst all independent travelers that gave a response to Overall Satisfaction and Gender, 41% are females who gave a Top 2 answer to Overall Satisfaction.
Scenario |
Total (qualified) |
Total Counts |
Base |
Percentages Calculation |
What does this mean? |
Percentage Base Total (Traveler Type: Independent) filter: Traveler type: independent |
Total (qualified)=566* The total is filtered by the selected element i.e Traveler Type: Independent | 230* The counts for Satisfaction and Gender are filtered by Independent Traveler. | 566* The base is the same as the qualified total for all cells. |
Traveler Type (Independent) percentages based on overall Satisfaction and Gender and Traveler Type (Independent) Total (qualified) filtered by: Traveler type: Independent |
Amongst Independent Travelers Qualified Total 41% of Independent travelers who gave a response to Gender and Overall Satisfaction are female Independent Travelers that gave a response to Top 2 Overall Satisfaction. Top 2 = |
Where to from here?
Learn more about Bases