array parsing (flatten, unnest), extractlist, unpivot, replacepatterns
This template flow shows you how to prepare and analyze survey data collected from a survey form. The survey form contains many different types of questions – such as single response, multiple choice checkbox and other combinations.
This template comes with an example survey (that’s built using Google Forms) where responses are collected in a CSV, but you can replace it with your own survey and response results collected in either CSV, Excel or Google spreadsheet (if using Google Cloud Dataprep).
The flow starts by doing some basic cleaning and renaming of survey question attributes in the Clean Headers recipe. It then forks into 4 separate recipes where each recipe processes one specific type of survey question. The recipes make extensive use of extractlist transform to collect responses into an array, and then various array parsing functions such as flatten to convert them into individual response rows. You can use this technique to work with many types of survey question responses.
Each of the recipes is annotated with detailed instructions, so feel free to step through them one at a time and customize it to your specific survey.
On Google Cloud Dataprep, this flow template is also available as part of a broader Smart analytics reference pattern.