This article describes how recode a numeric variable containing high or low outlier values:
And recode those values as missing data, so that these cases are excluded from calculations:
- A numeric variable containing high or low outlier values that you want to recode as missing values.
1. Select your numeric variable from the Data Sets tree.
2. OPTIONAL: Click the Values button from the object inspector to review the existing value attributes and determine the high or low values that you want to recode as missing data.
3. From the toolbar, select Anything > Data > Variables > Modify > Recode and then select one of the following depending on whether you want to set high or low values as missing:
- Set High Values to Missing in Numeric Variable(s).
- Set Low Values to Missing in Numeric Variable(s).
4. If you have multiple data sets, you'll next be prompted to select the data set containing the variable(s) that you want to recode.
5. A list of available numeric variables from the selected data set will be displayed. Select the variable(s) that you want to recode.
6. Enter the value that you want to use as the cutoff.
A group of tables showing the recoded question(s) will be added to the end of your report.