How To Remove a Category and Re-Base a Table
Often questions contain a response option that essentially means that the respondent has not provided useful data. For example, in a political poll if a respondent chooses the Don't know information then, from the perspective of trying to produce a prediction, the respondent has provided no data. In such instances the standard practice is to remove the Don't know data from the table and recompute the percentages (which is known as re-basing the data).
Single response questions
With single response data tables are re-based by setting the specific category as a missing value. For example:
- In Q the user right-clicks on a category and selects Remove which causes all tables using that data to automatically update with the category removed and the computations re-based.
- In DataCracker the user specifies to remove the category by selecting Data Manipulation > Data Values > Missing Values and then specifies the category to be set as Exclude from Analyses and then all analyses automatically update.
- In SPSS the process is a little more involved:
- Go to the Variable View.
- Click on the Values cell for the variable of interest.
- Specify the value(s) to be automatically removed and press OK.
- Re-run any analyses.
Multiple response questions
With multiple response questions there are two alternative ways of re-basing them:
- Compute new variables with missing values for all of the categories (this will not work if the data is set in a max-multi format).
- Filter the table.