Bulk declarations of variables in the codebook

Well-structured datasets usually follow consistent variable naming conventions. In such cases, these patterns can be used in the codebook to avoid overloading the dictionary and repeating the same declarations. A single-line pattern can declare a group of cognate variables, typically representing a multi-response, matrix question, or a block of standard questions.

1-tier Pattern

For example, the pattern below declares categorical variables for the brand dictionary. A single line replaces individual declarations for each brand in the list. This way, the brand list is declared only once on the Values sheet. You can then use the {brand} pattern to declare all variables that share the same set of options. This saves significant effort when your survey includes hundreds of brands.

When DataTile encounters a dictionary inside a pattern, it iterates over the dictionary's value labels. It uses value codes as variable names and injects value labels, midpoints, and other parameters from the dictionary when unboxing the pattern.

• NAME=UA_{brands} is unboxed as UA_1, UA_2, UA_3, UA_4, UA_5

• LABEL={brands} - Unaided awareness… will be unboxed as the list of corresponding labels.

2-tier Pattern – Matrix Question

Here’s an example of a matrix question. We are asking respondents about the packaging of the soft drinks they consume.

As before, all these variables can be declared with just a single line, assuming they follow a consistent naming pattern.

For example, having 100 brands and 7 package types, the pattern would define 700 variables in one line!