This article is part of the Definitive Guide to Uploading and Preparing Data.
Grouping is a very important first check to ensure that the datafile is set up correctly. In Crunch, you don't need or want individual variables ungrouped if they are meant to be together. It is important to read Recommendations for setting up a clean and duty dataset.
Very commonly SPSS files lack the metadata that defines the variable groupings. There are commands in SPSS syntax that define these groupings (MCGROUP and MDGROUP) a datafile (.sav). You have the option of setting this metadata before you upload the file to Crunch - in which case Crunch will read them in as 'real arrays'.
If the SPSS file does not group variables into arrays (categorical arrays, multiple response and numeric arrays) then you need to group them after you upload. When you make these groupings in Crunch, you derive arrays. This means you are making new variables and leaving the contributing variables intact. It is for that reason, the Crunch Automation commands have a handy argument (HIDE INPUTS) because you typically don't want the input variables hanging around (they are best stored away in the hidden folder. When they are hidden, they can still be used in other commands and calculations).
Note: if you use one of Crunch's importers (eg: from Decipher, Qualtrics, Survemoneky, etc) or from your own custom importer, or from a CSV + metadata description document (JSON)... then arrays are described on import already - and so you can skip this task. This is a good reason behind using direct importers as part of your Crunch workflow.
Crunch Automation
As making arrays is a common task (with SPSS files), so you may like to consider the Quick Editor for this purpose.
There are special Crunch Automation commands to deal with multiple response data that has not in a dichotomous format needed for multiple response variables. That's when CREATE MULTIPLE DICHOTOMY WITH RECODE is used.
In the web app
- You use the array builder from the + New Variable button (bottom left)
- The instructions for this, including video, are covered extensively here.
- Tip: As you don't need to see contributing variables to the array, you can hide variables.
R
- Use deriveArray()