In Methods to use in small projects - #2 by fnetX, I just suggested to reduce the usage of dropdowns as it is my personal opinion. Just an hour ago, I reviewed a change in a software that would IMHO benefit from less dropdowns. But I wondered what you think about it?
Do you have (or know of) any guidelines for things like dropdown usage that I can refer developers to?
they require one click to open in addition to the selection
users cannot see the inside while scanning a form
they are good when you have many options and expect few users to need to read/change them, you can conveniently hide the complexity behind an optional selection
they are bad when you have few options only, because then the space saving doesn’t pay off
they are bad when you expect most people to click through it, because then you double the clicks necessary for everyone
Depends on context, but generally agree; I would usually use them only if checkboxes and radio-buttons fail (cause these do not need to be activated to see the options).
In my opinion Vitaly Friedman is a reliable source as NN/G and worth to follow on LinkedIn. His referred article also contains a decision tree when to use which UI component.
Quite often it depends on the application, context, and design principles. We had a rule at one company that, the moment we hit more than 12 check boxes or radio buttons in a list it must be turned into a drop-down menu.
Of course there were other considerations that bypassed this rule such as clutter. We also don’t want a page of questions that makes the entire page filled with check boxes.
So I would suggest to think about the principles, perhaps take a look at everywhere there are check boxes and dropdowns today, and draft some guidelines around it.
I’m not really sure about your statement regarding the exception to the rule. Are you saying that clutter is an exception to the rule regarding checkboxes? In my humble opinion, clutter is precisely why such rules exist.
Regarding a whole page of checkboxes, I think the first questions should be:
Are we trying to solve a specific problem? Or are we trying to do too many things at once?
Which questions and options are meaningful and necessary to help the user complete their task?
Would a different information architecture be more useful and usable?
This is what distinguishes UX design from UI design: while UI design asks ‘How?’, UX design asks ‘Why and how?’
Actually a good exception to the rule may be filters on a table. There may be only 5 options, but you don’t want to have check boxes.
So yes, in essence it always comes down to clutter, but the rules really need to be defined generally and as a per pattern where exceptions exist.
The amount of times I have seen, when something is not explicitly said, a team of engineers you may not work with will take the rule literally and put check boxes everywhere. And by the time you or someone in your team learns about it, it’s too late and low priority to fix
