Outputs and Variable Names#

Each question produces one or more output values, and a fielded survey yields a dataset with a column per output. This page describes the Outputs page and the rules that name those variables.

The Outputs Page#

The Outputs page lists every output variable of the survey: the Variable name, the Source item with its detail, the Type, and the Rule that produced the name. A pencil icon marks names that were overridden in the question’s Outputs tab. The filter box matches variables, sources, types, and rules; Save as CSV… exports the table. Double-clicking a row opens the source item on the canvas.

Automatic Naming Rules#

Variable names derive from the question name according to the shape of the response.

A basic question produces one variable with the question’s name and the response’s data type.
A multiple choice question that allows more than one selection produces one variable per choice, named {prefix}_{choice}.
A choice grid with one selection per row produces one variable per row, named {prefix}_{row}.
A choice grid with multiple selections per row produces one variable per combination, named {prefix}_{row}_{choice}.
A question grid produces one variable per combination, named {prefix}_{row}_{column}.

In these patterns, {prefix} is the question name. {row} and {choice} are the category name of the row or choice when one is specified, and its code value otherwise. {column} is the column’s label with spaces removed when one is specified, and the column’s position otherwise. The General section of the Settings dialog can omit the question-name prefix from automatic names; see Settings Dialog.

Overridden Names#

A question’s Outputs tab can replace the automatic names with explicit ones, entered one per line in row order. The Outputs page and all exports use the overridden names.