This article is part of The Definitive Guide to Uploading and Preparing Data

Values are defined to produce means on tables. Labels are applied to categories so you know what the answer options are (eg: Male/Female).

As part of a dataset setup, you may need to change or remove the values.

• Remove the values if the means are nonsensical (for example, values on Male/Female don't produce meaningful means, so it may be useful to remove them)
• Put in class centers (midpoint recoding) for ordinal categories (such as age or income groups)
• Reversing or adjusting scales (eg: reversing 1 to 5 to be 5 to 1, or perhaps changing 1 to 11 to 0 to 10)

Likewise, the metadata on labels may be "messy" (for example, questionnaire or HTML code contaminates the labels in an SPSS datafile or is carried over into the subvariable labels on categorical array).

In the web app

• You can change values in the Variable Properties
  • Note: for Categorical Arrays you need to go into Edit for each sub-variable 

Using R

• See the Variables vignette
• For categories, use Categories() function and related functions - eg: names(categories(variable))
• For values, use values() and related functions - eg: values(categories(variable))

Crunch Automation

• SET NUMERIC VALUE command
• LABEL CATEGORIES command