I am using IBM SPSS Statistics. I have a string variable with a width of 8. However after I saved and reopened the file, the string width was now 24. What is happening?
string widths triple in size after saving data file
Starting with version 21, IBM® SPSS® Statistics operates in Unicode mode by default. Unicode is a type of character formatting that allows characters, including non-Western characters, to be displayed despite the native language ('locale') character set of the computer operating system where Statistics is installed. Because Unicode supports many different character sets ('locales'), more data may be required to support the display of native characters sets. This is the case with Western-language string data.Translating characters from non-Unicode ('locale') to Unicode results in as much as a tripling of the amount of information required to record string data,
Diagnosing the problem
Data files containing string data increase in size when opened in Statistics 21. The length of string variables increase in size in data files opened in Statistics 21. The changes in size persist when the data is accessed in other versions of Statistics.
Resolving the problem
Launch Statistics. Do not open any existing data file. Select the 'Edit' menu. Select the 'Options' menu. Select the 'General' tab. In the section labeled 'Character Encoding for Data and Syntax', select 'Locale's writing system'. Select the Apply' button. Select the 'OK' button. Open a data file using File->Open->Data, Select the check box for' Minimize string widths based on observed values' and open the data file. The extra widths will be removed from the string variables. For more information regarding Unicode, see the Help Topics (specifically 'Unicode mode').