I'm stumped by this one. Redesigning a large web app and one piece of the interface was this, where the user can select how they want to be notified by various things. I doctored the screenshot to keep the app private - really the list goes on for about a dozen rows. It's kinda ridiculous and I was just wondering if anyone has seen an alternative to this dilemma, because I haven't. The various things to be notified about have no correlation so you can't group them together. The only other thought I had was possibly converting the checkboxes into toggle switches. Perhaps there's a 3-point toggle switch that wouldn't suck?

double-checkbox list

  • There's a ton of stuff on checkbox alternatives in the archives. ux.stackexchange.com/search?q=alternative+to+checkboxes – dennislees Oct 6 '14 at 19:10
  • How about you have a few preset combinations (say three), which select appropriate options, then allow further customisation, only if the user needs it. – Steve Jones Oct 7 '14 at 9:38
  • @dennislees, I don't think that's a duplicate. It's dealing with as many as six columns of checkboxes, whereas this question is asking about two specifically. – Graham Herrli Oct 7 '14 at 14:57

The checkbox method meets user expectations

One of the most important things to do in any interface development is not confuse the user. Relying on established design patterns, such as checkbox for toggling independent settings, minimizes this risk. Given the purpose of the form, you should keep the checkbox interface because users will understand it.

You should continue to explore grouping

Your stated concern is that this list of events runs on too long, which could make it difficult for users to examine. In order to reduce the list to a more manageable size, you will have to group the list items somehow. Since you are having trouble coming up with groups, you should run a user test to find them experimentally.

  1. Write out each of the list items on an index card.
  2. Shuffle the cards up randomly.
  3. Give this deck to a representative user, and ask her to cluster related cards together. She doesn't need to name the clusters, and you should let her arrange the cards into as many or as few clusters as she deems fit.

You'll need to find three or four people to do this activity with, and make sure that they are familiar with any specific terms used in the list items. These people don't necessarily need to be current users of the application; in fact, it is better that they are not familiar with the current implementation of the application, because that could bias the results. Each activity will take a few minutes at most, and at the end you can look at all the clusters you found and neaten them up into categories.

By getting representative users to build your categories for you, you'll have a better chance of structuring the interface in a way that your actual users will find easy to work in.

| improve this answer | |
  • In this case user testing is basically impossible because I can't speak with the users. But I'm going to look a little deeper and see if there's a way to group some of them together. Thank you! – tylerl Oct 7 '14 at 15:03
  • It's not necessary to speak with actual users. You only need people who are similar to your users, and who will understand any specialized terms that are in your content. – Grubermensch Oct 7 '14 at 15:06

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