What branching logic can and can’t do

Branching logic shows or hides pages and questions based on a respondent’s earlier answers. It works by showing content to the people it’s meant for, not by jumping from page to page. Knowing this makes most rules easy to set up. For step-by-step instructions, see Show questions based on previous answers (Branching logic).

What branching can do

  • Show a page or a single question only to people who gave a certain answer.
  • Combine several conditions in one rule, for example students who also live on campus.
  • Give different groups different paths through one survey, while everyone still sees the shared pages, such as background questions at the end.

How rules behave

  • Showing something to one group hides it from everyone else. If a page is set to show for people who answered “Student”, nobody else sees it.
  • One rule per page. A rule applies only to its own target. If four pages are for students only, add a rule for each page; use Duplicate rule to save time.
  • Think in terms of what to show. Rules are easiest to follow when you set what each group should see, rather than what they shouldn’t.
  • Build the survey first. Add branching rules once all pages and questions are in place.

What branching can’t do

  • Loop back to an earlier page. The survey always moves forward in page order.
  • Send respondents to a specific page. Rules show or hide content; they don’t jump. To end the survey early for people who don’t qualify, place a page explaining why just before the last page, and show all other pages only to people who qualify.
  • Shuffle the order of pages.

Want to shuffle answer options? That isn’t part of branching: turn on randomized order in the question’s own settings.

Test your rules

Go through the survey in Preview once for each group of respondents. You can use the navigation tool in the bottom-right corner of Preview to skip past required questions. Then publish for testing and try it on a phone.

One survey or several?

Use one survey with branching rather than separate surveys for different groups. Everyone’s answers then stay in one dataset, and you can compare groups with filters in the analysis tool. See Keep map and standard questions in one survey.

Was this article helpful?
0 out of 0 found this helpful