If you have questions that have missing values, either because they are missing in the original data file or because they have been recoded as missing (e.g., by right-clicking on a category and selecting Remove), they can be replaced with an average value.
Method
- Check that the Structure of the variable sets is either Numeric or Numeric - Multi (and change it if necessary).
- Create a table with the variable set in the rows by dragging it onto your page to display as a summary table.
- Export the table to Excel via Share > Export Report > Excel.
- Go to the Data Sources tree.
- Select each variable in the set, one at a time, and:
- Make sure to have only the one variable selected, and from Properties
, click Data > Attributes > Values.
- Set the missing data row for the categories that are being recoded as the average to Include in Analyses in the Missing Values column.
- In the Value field for the category that represents the missing data, paste in the value from Excel of the average.
- Press OK.
- Make sure to have only the one variable selected, and from Properties
Please note that:
- Recoding missing data as averages prior to performing multivariate analyses, such as cluster analysis, is not generally considered to be good practice.
- Latent Class Analysis automatically deals with any missing data.
Next
How to Turn Off Missing Data Selection for Specific Values
How to Set Value Attributes for a Binary-Multi and Binary-Grid