Easily create and recombine categorical arrays and multiple response variables


Users can now combine variables with common categories to analyze together as a group, and summarize arrays by choosing categories of interest.

Jul 16, 2020

Array variables allow you to analyze together a set of variables that share a common structure and theme. For example, surveys often ask respondents to use a common scale to rate or respond to different prompts, sometimes in a grid, all sharing a common set of response categories. When analyzing the data, a categorical array variable lets users see all the items together. This structure allows users to easily compare lists of brands against a common set of attributes, or compare a set of statements against a common agreement scale — something that is much harder to do with distinct categorical variables. Categorical arrays are displayed as two-dimensional tables or stacked graphs by default in Crunch, letting you compare the complete distribution between items.

A multiple response variable is a kind of summary of the items or input variables in a categorical array. In its simplest form, responses take the form of either “selected” or “not selected”. Crunch uses this form (called “multiple dichotomies”) to represent survey questions where respondents can select multiple answers from a list of responses (“Choose all that apply”) — each response in the data is a separate variable, but for analysis, they need to be grouped together into a multiple response variable.

Multiple response variables in Crunch allow you to choose more than one selected category, so you can also summarize an array of “level of agreement” into a multiple response view of the percentage who selected either “Agree” or “Strongly agree”. Multiple response variables are displayed as a ‘flattened’ form of a categorical array, allowing you to analyze this more condensed form, crossing it by demographics or trending over time, etc. These summary variables are sometimes called “top 2 boxes” referring to selecting the subtotal of the “top two” categories on a scale.

The following videos demonstrate how to create these new variables. First, we create a new array from a group of categorical variables that have a common set of answer categories.

Then, we create a simplified multiple response variable from a subset of those items that we are interested in.

For more information see the help center.

We’d love your feedback at support@crunch.io.