How can I Use Sociometrics datasets with other statistical packages?
Sociometrics datasets include program files written for
the SPSS statistical package, along with "generic" raw data
files. Other statistical packages may be used to read in and format
raw data files. One way to do this is to edit an SPSS program file so
that it contains the appropriate variable information and program syntax
for the alternate statistical package. The raw data and SPSS program
files are
organized as follows:
- All raw data files are named by Data Set Number(s) and the file extension ".RAW".
- All raw data files contain data values in "fixed-column" positions. Data values for a particular variable always appear in the same column positions throughout a raw data file.
- Most raw data files have been formatted with a standard logical record length of 80 characters. For SPSS-X and older data set files, the FILE HANDLE command in the SPSS syntax program indicates the logical record lengths in the LRECL subcommand.
- No "delimiters" (such as commas or blank spaces) are used to separate neighboring data values on records in the raw data files. This is typical of fixed-column data files.
- Many raw data files employ multiple records per case, (i.e., a block of several lines is required to hold all the data values for one study respondent, or one case).
To read in and format raw data file with a statistical package other than SPSS, you will need to know at least four things: (1) the number of records per case (i.e., number of lines of raw data employed to hold all data values for a single respondent); (2) the logical record length of the raw data file; (3) the column positions of the records in the raw data file that correspond to each variable of interest; and (4) the variable names. This information is contained in the DATA LIST section of the complete SPSS syntax program files on the CD-ROM or downloaded dataset.
SPSS syntax files are named with the .SPX (for Unix/mainframe) and .SPW for SPSS/Windows. Older datasets may use SPSS extensions of SPC for personal computer.
To read the raw data file for any Data Set into another statistical package, copy and edit the FILE HANDLE (if required) and DATA LIST sections from the SPSS program file of interest so that the contained information corresponds to the requirements of the alternate statistical package.
You may also be able to copy and edit other sections of an SPSS program file for use with the alternate statistical package to further document and format the data values:
- The VARIABLE LABELS section contains a list of all variable names and variable labels.
- The VALUE LABELS section contains a list of variable names and their corresponding response options.
- The RECODE section contains a list of variable names and any data values that need to be changed (some SPSS program files do not include RECODE commands).
- The MISSING VALUES section contains a list of variable names and data values to be specified as missing data (some SPSS program files do not include MISSING VALUES commands).
- The FORMATS section contains a list of variable names and required formats. For example, an "income" variable with decimal values such as 38000.55 might be assigned a format of "F8.2", meaning a total format of eight places with two decimal places (some SPSS program files do not include FORMATS commands).
