Training

You likely receive datasets from a variety of different sources, including the IBM SPSS Data Collection+ data file format (*.ddf or *.dzf), which has the rich metadata stored in an IBM SPSS Data Collection metadata document (*.mdd). You may also work with data files in formats that don’t include a separate metadata document file, such as the IBM SPSS Statistics++ file (*.sav) format.

One difference in using these datasets is the existence of grid variables in the IBM SPSS Data Collection datasets. These loop-type questions include a shared category list – usually a rating scale – that is repeated for each of several iterators or sub-questions. The grid structure includes one variable for each sub-question.

When viewing IBM SPSS Statistics files from within IBM SPSS Reports for Surveys, sub-questions appear as individual variables. However, if you’re using an IBM SPSS Data Collection dataset, sub-questions appear under an expandable structure in the variable listing.

The advantages of using grid variables include the ability to easily set up summary “grid tables” and to make scale changes across the sub-questions. In this tip, you’ll learn how to create new grid variables from a non-IBM SPSS Data Collection dataset.

Create a new grid variable

IBM SPSS Reports for Surveys 5.6 and IBM SPSS Reports for Surveys (Server) 5.6 provide an option to create new grid variables. This capability assumes that you select variables with a common scale. However, you should proceed with caution: the software will allow you to create grid variables from variables with non-matching scales, adding all categories to the shared category list.

To create a grid variable from other variables in the dataset:

• Select the variables from the Variables pane. Use the Shift or Ctrl keys to choose multiple variables.
• From the menu, select Variables | New | Grid
• In the New Variable dialog box, name your new variable and make any other changes to the sub-question list
• Click on Save and Close when you’ve completed the edits. The new grid variable will appear beneath the other variables in the Variable pane.

When you expand the grid variable, you’ll see the main grid name, ServRating, the shared category list, Column and each subquestion, ServRating_Cart_Ratings, ServRating_Order_Ratings, etc.

Note: Each sub-question variable will match the name of the individual variable used to create the grid variable.

Customize the new grid variable

If you’d like to change the name of the shared category list, Column, take the following steps:

• Edit Column, and save the list as a new variable, Ratings
• Delete Column
• Select any numeric variable to weight a data table

You’ve created a new customized grid variable:

Conduct a more efficient analysis with the new grid variable by:

• Dragging-and-dropping the grid variable to Add Grid/Loop to create a summary table
• Editing the shared category list, Ratings, to add nets or modify categories
• Using the grid variable with Tables | New | Table for Each Selected Variable to create a table for each slice

+IBM SPSS Reports for Surveys and IBM SPSS Data Collection were formerly known as PASW® Reports for Surveys and PASW® Data Collection.

++IBM SPSS Statistics was formerly known as PASW® Statistics.