Exporting data from CL

This is the facility of transferring data entered with CL into another software package using Triple-S or to a specialised statistical program, for example SPSS or SAS.

To export data, from the "Main" window, select [Data] [Export data].

In most instances you can simply choose the type of export that you, or your clients' require. However, if the data file is very large it may be too big for the other program to use, or import. If this is the case, you will need to generate and amend a .DTV file to control which, and how much, information is exported. You would also need to do this if you did not wish to export all of the variables you had created, or if you had a syndicated data file (where each member of the syndicate required different data from the file). A .DTV file should also be created if you have exported the data file before and you need to match the data previously exported. For more information see section DTV generation and use

For information about the selection [Reflect (QRF)] see section Creating files for use with Reflect

  • Selecting [Triple S XML (XML, ASC)] will produce a data file with the extension .ASC and a variable file (.XML) suitable for use with market research programs capable to reading data in Triple-S XML format. A .MAP file is also produced.
  • Selecting [Triple S (SSS, ASC)] will produce a data file with the extension .ASC and a variable file (.SSS) suitable for use with market research programs capable to reading data in Triple S format. Triple-S 1.1 (with locations) and 1.0 (without locations) are available. A .MAP file is also produced.
  • Selecting [SPSS Windows (SPS, TXT)] allows you to export data to the statistics package SPSS for Windows. Two files are generated from the SPSS for Windows export; a syntax file which holds the variable definitions, with the suffix .SPS and the data file with the suffix .TXT. To import the data file into SPSS for Windows, you can either double click the .SPS file (in the File Manager or the Windows Explorer) or once in SPSS you open an SPSS Syntax File and then select [Edit] [Select all] [Run]. Note that if the SPSS output files are to be used on a different computer (not the machine used to export the data) you, or your client, will need to modify the .SPS syntax file, so that the correct path for the .TXT file is shown.
  • To export data suitable for use with the SAS statistical program, from the "Export data" dialog box, select [SAS (SAS, TXT)].

For SPSS and SAS exports response texts beginning <?> are not output.

There are two methods for outputting multi-coded variables. A new column can be used for each response (0 or 1) or a spreadfield can be requested. A spreadfield is a set of single-coded variables that contain the answers to a multi-coded variable.

  • The [Quantum (RUN, ASC)] data export selection outputs the data in a format suitable for use with the Quantum analysis package. The output data file has the suffix .ASC. Three .RUN files, which hold question and variable information, are also generated with the Quantum export, these are MEAN.RUN, BASE.RUN and ACARD.RUN, and these files should be used with the .ASC file, to ensure that questions and variables need not be redefined.
  • The [Comma delimited (CSV)] produces a comma separated variable data file with the extension .CSV . In the file the data for each question on your questionnaire is separated by a comma. This type of exported data is suitable for direct import into most spreadsheet and graphics programs. In the standard .CSV export the data from multi-coded questions is output as a series of 0 and 1 codes (a 1 denoting that a response was selected). Each 0 or 1 is held in a separate field (so that one number would be held in each cell of the spreadsheet). It is also possible to select [One column for multi-coded] before exporting the .CSV file, which will cause all multi-coded responses ( 0 and 1 numbers) for the same question, to be held in the one field (spreadsheet cell). See also the LCSV (list CSV) command.
  • Selecting [Fixed format ASCII (ASC)] exports CL data as an ASCII file, with the extension .ASC. Each record is held as one long line of data. Most modern analysis programs can accept this form of data.
  • Selecting [Fixed ASCII in 80 column lines (ASC)] will export the CL data as an ASCII file, with the extension .ASC, but with only 80 characters per line (record) of data. This type of file is not often used, except by programs which are not able to accept data consisting of more than 80 characters per line.

If the [Max sizes for numbers] box is selected any integer questions will be exported holding the maximum number of data locations (9) regardless of the largest number held within the question. Normally, when CL exports a file it scans the integer questions and exports only enough data locations to hold the largest number entered for the question.

IMPORTANT: If you intend to output a number of waves of data for the same project (for example, a new data file is exported each month) you must select [Max sizes for numbers]. This is because in wave 1 the largest number entered for an integer type question might be 999 - so CL would use 3 data locations to export the data; in wave 2 the same question might hold a number of 1000 - in which case CL would use 4 data locations and the exported data would not match that exported in Wave 1.

With [Max sizes for numbers] selected, single-coded questions will use the data location width needed to hold the longest response number; multi-coded coded and open ended questions will have one response in each location (so if the multi-coded question had 8 responses, a data location width of 8 would be allocated); character questions use the width as specified in the set-up, integer questions are allocated a width of 9 data locations and float questions are allocated a width of 20.