All commonly used statistics on tables, such as %, Column %, and Average, are computed using the base. It is often necessary to rebase tables (i.e., modifying the sample size used in the calculation of statistics). For example, tables may be rebased to address data integrity issues or to see results among a particular sub-group. This can be done using a range of different methods, as follows.
- Filtering the table
- Modifying the Value Attributes
- Creating new variables that are filtered
- Deleting cases
Filtering the table
The most straightforward approach to rebasing a table is to create and apply a filter. Applying a filter to the table will only change the data shown in that specific table, and will not change the data for the variable set used throughout the document. To do this, you will need to use one of the other methods below.
Modifying the Value Attributes
A variable set's Value Attributes determine which categories are included and excluded from tables and analyses (and therefore the base) in the Document. Here you can increase or decrease your base by choosing which categories and values to include and exclude from the base. For example, if you are looking to increase the base of your table, like rebasing to the Total Sample, you will need to change the Value Attributes for the Missing Data category (and possibly others) to Include in Analyses. Another example of when to change the Value Attributes is when you want to remove respondents who selected a particular category (such as None of the Above or Don't know) from a categorical (Nominal) variable set.
To do so:
- From the Data Sources tree, select your variable set.
- In the object inspector, click Data > Properties > Values.
- Adjust the Missing Values field for the category you want to remove from the table to Exclude from analyses.
A shortcut to exclude categories is to right-click on their row or column headings on a table and select Delete. This has the same effect as changing the category to Exclude from analyses in the Value Attributes.
Changing which categories have a Value of NaN does the same for Numeric variable sets. You can also choose which statistics (percentages or averages) should include this category if needed, see Value Attributes.
Rebasing Binary-Multi variable sets can be a bit more involved depending on how the data for the multi-response is coded, see How to Set Value Attributes for a Binary-Multi and Binary-Grid. To rebase a Binary-Multi based on categories shown in the variable set or another see the section below.
Creating new variables that are filtered
There are automations that can create copies of your variable sets that are rebased. These can then be used to create rebased tables.
Rebase to NET
You can rebase a table with a Binary variable set based on respondents who selected certain options (such as a subset of brands) or remove respondents who selected certain categories (like a None of the Above or Don't know option).
- Optional: If removing respondents for an option from the base, right click on the variable(s) for that option in the Data Sources pane and select Split to remove it from the set.
- Rebase the table based on respondents who selected at least one of the remaining options. From the Data Sources pane, select your variable set.
- Click + > Data Quality > Rebase Multiple Response Data in Variable(s) to NET (see How to Rebase Multiple Response Data in Variable(s) to NET for more details).
Use responses in one variable set to filter another
You can also create a new variable set that is rebased used respondent's response from another variable set (you can think of this as rebasing one table based on another). For example, you can rebase a variable set for Brand Consideration for each brand by only those respondents who marked that they were aware of the specific brand in a Brand Awareness variable set. From the Data Sources pane, select a variable set you wish to filter. Then hover and click + > Filter > Filter One Variable Set by Another. You can select the variable set to rebase and the variable set to use to rebase, see How to Rebase One Question Based on Another Question for more details.
Deleting cases
Where there is a desire to rebase all the data in the project by removing some respondents, this is best achieved by deleting cases.