I've been handed a settings page to review. The settings mix true/false and open-ended items. Some of the open-ended items can be long and wordy. This is an example of the current design:

Options list with X icons and open-ended attributes mixed together

Mixing the booleans with the long-form attributes seems wrong, but I can't recall any principles to explain why. It looks cluttered and messy when there are dozens of them. Moving the long-form questions away from the booleans seems much cleaner.

Same settings but with the true-false questions clustered together

I am going to make a case to get rid of the View state altogether so users don't have to enter an edit mode to update settings, which should help - but meanwhile, why doesn't mixing booleans and other datatypes "work" visually? What principles might explain that?

2 Answers 2


I think any number of the Gestalt Principles can be used to explain why the combination of things might not look 'visually appealing'.

  • Similarity: the content/data type of string is not the same as boolean elements
  • Continuation: the alignment of the different amount of content creates a disruption to the alignment of other content
  • Symmetry and Order: there is an imbalance between the number of string elements compared to boolean elements

Any real task is not 'settings mixing' but it is 'settings grouping'. Grouping doesn't take in mind visual appealing as the main goal. Good user experiense is the main goal.

Suppose you have options which are cross-dependant. If you change 'Allow Option Two' in your form and 'Type of Option Choosen' become enabled is it usable to jump somewhere to edit it and then jump back to proceed with next setting whish is 'Allow Option Three'?

If possible you may use another techniques to achieve your goal. But this depends on your real options.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.