(I have first noticed this with MS Office applications, but I have seen it also ...somewhere in Gnome, I can't remember where...)

The idea

If a dialog pops up that basically asks you to choose only one option out of single radio button sets (e.g. "Send Response / Do not send response / Edit response before sending"), you can double-click the radio button of your option to dismiss the dialog with that option. Here the double-click works as a macro for choosing option and clicking "OK".

I like this idea because this serves as a compromise between having only buttons (which you can click accidentally) and having traditional radio buttons (which in fact you still have). If applied to certain class of dialogs (i.e. not a critical dialog, nothing but one fixed group of options + OK/Cancel), it can significantly improve user experience.

I wish I would see more applications adopt this behavior (just for the class of dialogs, of course), but before posting enhancements:

Does this behavior have any significant pitfalls or drawbacks? Can it become annoying or confusing? Other ideas?

The benefit

A nice example of a dialog where this is implemented can be seen on Figure B in this article on TechRepublic.
excel paste special dialog

The typical scenario when using this dialog is when you manipulate data, formulas and styles around, trying to figure out the best layout. If you are not a newbie in Excel, you know what you want to choose already when selecting data.

So being able to dismiss this dialog quickly can significantly decrease chance of losing concentration on the content. There is literally half of operations that require targeting and moving the mouse: instead of clicking the option and clicking OK, you just double-click the option. (You will feel the difference when designing tables.)

Yet the list of options can still be quite long, but well ordered.

Meta-edit: I'm really sorry if I give the impression that I tend to refute everything. I do appreciate every single idea posted or commented here (I haven't down-voted anything here yet). I was planning to give all of you a big thanks as soon as I find the answer, but it's fact that I owe it to you already. I either simply do not agree with everything -or- It does not answer my question. Maybe it's problem with my definition of the question, for which I apologize.

  • 1
    Two thoughts. 1) There's nothing about the control that lets a user know there's more behavior to be had...currently it's not convention, so without some sort of PSA to inform everyone, it's unlikely to catch on. If the radio button changes how it's displayed to look more like a confirm button after it's selected, that would probably be awesome. 2) What if the system is lagging, and you click multiple times (as frustrated users often do) trying to get a response? What if it's something important like Save/Discard? – Brad Orego Jan 18 '12 at 19:40
  • Can you post wireframes or screenshots? – dnbrv Jan 18 '12 at 19:51
  • @dinko628 1) I don't think it's necessary. (However, one can always decide to implement some kind of Mr. Clippy :)) Changing appearance would bring a risk for users that are afraid of new things. 2) Good point, but extreme case, and that problem is not specific only to this concept. (You might as well accidentally launch something or answer two consecutive dialogs.) If there is a risk, it's up to designer to consider using radios or separate buttons. (I have edited the Q to better specify the class) – Alois Mahdal Jan 18 '12 at 20:16
  • @dnbrv Not directly I think--with reputation of 1 :) But there's nothing here that screenshot could clarify--it's the same old radio group just with double-click enabled. – Alois Mahdal Jan 18 '12 at 20:20
  • 1
    @AloisMahdal What exactly is the benefit of making this control a radio button, then? It seems to be something that would be better suited as just a button. Radio buttons tend to serve well for mutually-exclusive, non-actionable items (e.g. settings). If it's actionable, I don't see how radio buttons offer benefits regular buttons don't – Brad Orego Jan 18 '12 at 20:31

What are the pros and cons?

  • If the user knows about this feature then they gain a tiny advantage in task-time.

  • If the user doesn't know about this feature then very bad things might happen (an accidental double-click resulting in action being taken).

Given that (a) it's changing a long established UI convention, (b) there is no clue or fore-warning of this affordance, and (c) the potential losses outweigh the potential gains ... it's a bad idea.

For the example given, I'd also note this: I'm no Excel newbie, but when I use the Paste Special dialog I drive it using the keyboard. Cmd-Shift-V t Enter I'd bet that many other Excel non-newbies would also do the same.

Which raises another point - If this double-clicking radio-button is meant to be an aid to more experienced users, then it is competing for mind-space with the well established mechanism of using keyboard shortcuts. But only for radio buttons. So the user would still need to learn the other forms of shortcuts (I see MS hasn't implemented double-click-submit on checkboxes. Of course.)

I really don't see an upside to this at all.

| improve this answer | |
  • (a) yep. it is. But in constructive way. It does not change any documented behavior (plus I bet that many people already use other GUI that has already implemented this long ago and did anyone complain?) (b) not a bad point, but it's rather question of polishing the actual implementation (c) only if it's misunderstood but that does not mean it's a bad idea – Alois Mahdal Jan 19 '12 at 2:03
  • And about the competing: No, never. (Or maybe in very bad GUIs.) Mouse is not here to compete with keyboard. Instead, good GUI should always be comfortable with both modes: when keyboard is primary source or when mouse is. (Or even if one is missing, or not usable for some people.) How does this feature compete with the first mode? No way. It only enhances the seond one. – Alois Mahdal Jan 19 '12 at 2:14
  • 1
    And please let's avoid the risk of turning this into Holy Keyboard-Mouse war. :) – Alois Mahdal Jan 19 '12 at 2:16
  • 1
    @AloisMahdal Why do you ask for advice when all you're going to do is refute everything that the people giving you advice tell you? It seems like you've already made up your mind on the issue, and you're just looking for validation :/ – Brad Orego Jan 19 '12 at 6:01
  • @dinko628 Sorry, hopefully my edit will explain. – Alois Mahdal Jan 19 '12 at 12:32

My only concern with double clicking on radio buttons - it is not used a lot, and people might not understand that they need to double click without instructions, especially in online/web applications where double click is not that common.

| improve this answer | |
  • Isn't this exactly what the question is asking...? This seems more like a thought than any sort of answer to me... – Brad Orego Jan 18 '12 at 20:23
  • For online based applications / websites it is not safe. For software, it will be safer, but still not good idea, since this technique is not common and people will have hard time figuring this out without separate instruction. – Michael Jan 19 '12 at 16:50
  • people don't need to double click. They can double click if they want. But they can also click the OK button (or press Enter). This is just an additional option to submit the dialog. – splintor Jan 15 '17 at 18:05

I don't see the point. If the intent is to pick an option and close, why not make it a simple two-button interaction, both of which will set the value, then close the window. This is akin to how JS confirm windows work.

| improve this answer | |
  • Because radio buttons will give you something that action buttons can't, while still avoiding the messy look: 1) the text can be much longer without getting ugly--can even have variable length (e.g. include a human name or address) 2) the same applies for number of options 3) you can easily support one or more choices with additional widgets like check-box or a small text fields. Not all options must necessarily be meant to be used in the "click and dismiss" nature – Alois Mahdal Jan 18 '12 at 20:59
  • 4) you might as well have a dynamic part of the dialog (e.g. a tiny preview) which would change based on selected option – Alois Mahdal Jan 18 '12 at 21:13
  • Hmm...not sure I'm following you. I'd also suggest posting some diagrams/wireframes/examples if you can. – DA01 Jan 18 '12 at 21:23
  • I have elaborated on the question, hopefully making it more clear. Screenshot linked :) – Alois Mahdal Jan 18 '12 at 21:47
  • Ah...your updated question does clarify things. It does seem to have a purpose there (though I'd never encourage anyone to copy MS Office's bloated dialog windows to begin with). You are asking if it's "safe". It depends on what the action does, but in that context, it's certainly safe...just perhaps annoying for people that don't know about it and are habitual double-clickers to begin with. – DA01 Jan 18 '12 at 23:58

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